LIBRARY

OF THE

UNIVERSITY OF CALIFORNIA.

Deceived APR 1$ W93 , 189
Accessions No. 5..Q ^\ b..Q Class No.

Library

THE STORAGE

-OF—

ELECTRICAL ENERGY

ENTERED AT STATIONERS HALT,.

0? THB

TJIIVBRSITT

/-0^~^^-v £

THE

STORAGE OF ELECTRICAL ENERGY

AND RESEARCHES IN THE

EFFECTS CREATED BY CURRENTS COMBINING QUANTITY
WITH HIGH TENSION,

GASTON PLANTE,

LICENCIE ES-SCIENCES PHYSIQUES,
AXCIEN PROFESSEUR DE PHYSIQUE A L'ASSOCIATION POLYTECHMQUE

ETC.,

FROM

1859 TO 1879.

WITH EIGHTY-NINE ILLUSTRATIONS.

TRANSLATED FROM THE FRENCH BY

PAUL BEDFORD ELWELL,

(OF ELWELL-PARKER, LD.)

LONDON :

WHITTAKER & CO.,

2, WHITE HART STREET, PATERNOSTER SQUARE, E.C

1887

?

Engineering
Library

A Sa Majestt

DON PEDRO D'ALCANTARA
Empereur du Brdsil,

Associe Stranger de I' Academic des sciences de
£ Institut de France.

Sire,

Je prie Votre Majestd de daigner agre"er la
de'dicacc de ce Livre comme un faible tdmoignage
de ma profonde reconnaissance.

Vous avez /// le premier a encourager mes
travaux. Apres avoir assist e1, en 1872, a mes
experiences sur les courants secondaires, au Con-
servatoire des Arts-et- Metiers, Votre Majestd a
bien voulu, en 1877, honor er deitx fois de sa
visite mon laboratoire de la rue de la Cerisaie,
dans ce qiiartier du vieux Paris, ou les noms
d Henri IV et de Sidly sont encore vivants.

Cent soixante ans auparavant, Pierre le Grand
venait y habit er Photel Lesdiguieres, et, vers la
fin du dernier siecle, Pillustre Franklin assistait,
dans Chopital des Ce'lestins, a des experiences
a* tlectricite1.

La presence, dans ces m£mes lieux, de Votre
Majest^ poursuivant son enquete sur tons les
pr ogres utiles a I 'humanitk, aj out era une nouvelle
page a nos anciennes traditions et un pr^cieux
souvenir qui sera conserve1.

Je suis,

avec le plus pro fond respect,

Sire,
de Votre Majeste,

le tres-humble et tres-obttssant serviteur,
G ASTON PL ANTE.

ERRATA.

PAGE 14 — Line n from bottom, after "like" insert "the oxides of."

„ 18 — — 5, after "two cells" insert "acting on a voltameter with
wires."

it 22 5, for "Dewar" read "Dumas."

11 48 — — 9, for "production" read "reduction."

M 53— —14, for "Bollot" read "Boillot."

ii 53 — — 16, for "positive" read "negative."

•I 53— — I9> for "production" read "reduction.'"

•I 69 — — 5, for "also used in connection" read "compared."

ii 69 — — 8 from bottom, for "powerful" read "potential."

M 75 — — 4, for "Thomson" read "Thomsen. "

H HI — — 12, for "Willigan" read "Willigen."

M 126— —7, for "Sheath" read "Sheaf."

ii 37 — — 3 from bottom, for "combination of lime and silica"
read "lime combined with silica."

ii 138 — — 3 from bottom, for "silica" read "silicium."

H 141 — — 2 of the note, for "p. 58" read "p. 44"

ii 148 — — 7 of the note, for "rorid" read "roric."

M 186 — — 1 6, for "Brissou" read "Brisson."

H 209 — — 11^ f°r "supplied by" read "in connection with."

,. 221 — — 4 from bottom, for "fig. 80" read "fig. 79."

„ 223 — —13, for "fig. 79" read "fig. 80."

.. 238 — — 6 of the note, for "Vol. i" read "p. 214."

ii 257 ii, read "when there is a break or change in the

material."

PREFACE.

This work includes the main results of the researches we
have contributed to the Academic des Sciences, or published
in various scientific^periodicals, from the year 1859 to 1879.

It is divided into six parts.

The first comprises a description of the experiments and
apparatus we have made known for accumulating or transforming
the energy of the voltaic battery by means of secondary currents.

The second part contains an enumeration of the applications
to which these experiments have been put, and of others which
might be carried out.

The third part relates to phenomena observed with electric
currents of high tension obtained by the means described in the
first part.

The fourth part treats of analogies which these effects seem to
present with many great natural phenomena, and the inferences
drawn from them in order to explain these phenomena.

The fifth part embraces the description and study of the effects
of a new apparatus, by means of which we have tried to
transform, in the most complete manner possible, dynamic into
static electricity, and which we have distinguished by the name of
rheostatic machine.

The sixth is devoted to a concise enumeration of analogies
which the electrical phenomena (particularly those which we have
noticed with currents of very high tension) present with effects
produced by mechanical actions, and to the inferences drawn
from them as to the nature of electricity.

The reader who will only admit absolute deductions from
facts, may omit the fourth part, in which induction has
considerable play.

Nevertheless, we thought we ought not to pass over in silence
some of the ideas which the results of our experiments presented,
and the real or apparent analogy which they show with natural
phenomena.

Authors are often blamed for not having understood the
inference to be drawn from facts which they have observed, or for
not having perceived the theoretic sequence, and the applications
to which it may lead. We have sought to be free from this fault,
which, besides, has not always foundation ; for it is rare for
anyone who has patiently studied nature, seeking new facts, not
to also meditate upon their significance, and, as nature is seen
complete in each one of her manifestations, it is difficult for the
student to avoid being led to generalise the result of his observ-
ations. This tendency becomes, no doubt, another reef upon
which one may strike ; yet, science would not lose, we think, by
these generalisations, or by the hypotheses to which they lead,
from the moment they rest no longer upon pure imagination, but

are inspired by attentive observation of facts, and are set forth,
besides, with reserve, without elevating them as doctrine, without
affirming that they are true.

This is what we have tried to do, and, to follow the example of
one of the greatest thinkers of past ages, we will humbly say, in
publishing these researches : " Queer o, Pater, non affirmo"(t)

(i) St. Augustine.

FIRST PART.

The Accumulation and Transformation
of the energy of the Voltaic Battery
by means of Secondary Currents.

CHAPTER I.

Secondary Currents. — Voltaic Polarisation. — Study

of the Secondary Currents produced by different

Voltameters. — Conclusions.

1. SECONDARY CURRENTS were observed at the beginning
of this century, shortly after the discovery of the Voltaic Battery.
Gautherot, a French Scientist, was the first to discover, in i8oi,(l)
that platinum or silver wires which had been used to decompose
saline water by this battery, possessed the property, after having
been cut off from the battery itself, of giving an electric
current of short duration.

Ritter(2) made the same discovery at lena, with gold wire, and
made the first secondary battery, by superposing a series of pieces

(i) Memoirs of the Learned and Literary Societies of the French Republic, 1801.

(2) See Exposes des Travaux de Ritter par CErsted. Journal de Physique

1803. t. LVII, P. 345-

2

of gold, separated by cloth discs, moistened with a saline
solution. This battery, inactive in itself, after having been
submitted to the action of a voltaic battery of a greater number
of cells than that of which it was composed, could give for some
moments a current in the opposite direction to that of the voltaic
battery. This current took the name of secondary current.

Ritter varied the kind of metal, and the number and surface of
the plates composing the secondary battery.

He used platinum, copper, brass, iron, and bismuth, and found
that gold, platinum, and silver, gave a stronger secondary current
than any of the other metals. He noticed that carburet of
iron, and peroxide of manganese, gave still more marked results ;
but he obtained no effect with lead, tin, and zinc.(l)

The Secondary Batteries that Ritter employed more especially
in his experiments, were formed of discs of copper separated
by circular pieces of cloth, moistened in saline water, or sal
amoniac. By charging a secondary battery formed of a column of
fifty copper discs, by means of a zinc and copper voltaic battery,
of a hundred pairs, he obtained decomposition of water, various
chemical or physiological actions, and, in general, all the effects
produced by ordinary batteries. Nevertheless Ritter's Secondary
Batteries, in consequence of the objections arising from their
arrangement in a columnar form, or crown of cups, like that given
to the voltaic battery itself at that period, in consequence also of
the brief duration of the currents produced, and the necessity
of employing a battery of a greater number of elements to charge

(i) We shall explain further on (22) why Ritter obtained no result with lead, the
metal which we have on the contrary exclusively used for obtaining powerful Secondary
Currents.

them, could not be advantageously applied either in scientific
research or commerce.

2. The secondary current produced by one of Ritter's
batteries has been explained by Volta, Marianini, and Becquerel,
who have shown that this current arose from the formation of
acid and basic deposits upon the metal discs, in consequence of
decomposition of the salt impregnating the moistened pieces of
cloths, under the influence of the primary current. Becquerel in
particular proved clearly the production of an electric current by
the reciprocal action of an acid and its base, by surrounding two
platinum plates, one with an acid, the other with a basic solution,
united by a moist conducting material. He thus formed a battery
element of which the plate immersed in the acid solution,
constituted the positive pole, and the plate in the basic solution
formed the negative pole. The direction of the current of the
secondary battery became thus explained, this direction being
such that the disc in connection with the positive pole of the
primary battery, is itself the positive pole of the secondary battery.

3. About 1826, attention was again called by de la Rive
to the secondary current arising from platinum plates in a
voltameter filled, not with a salt solution, but with water slightly
acidulated by sulphuric acid, or even with distilled water.

As the presence of acid or basic deposits was out of the
question in this case, the secondary current was at first attributed
simply to physical effect produced by the primary current, to a
special polarization of the plates under its influence, and this
current received the name of Polarisation Current.

4. This term has remained in scientific use ; but it has since
been found that the gases developed around the plates, even in a

very small quantity, must be the cause of the current; for
Matteucci obtained a current with platinum plates, previously
immersed in oxygen and hydrogen gas, and Mr. Grove made a gas
battery, by coupling up a certain number of elements formed of
platinum plates, immersed at one and the same time in acidu-
lated water, and inverted test tubes of oxygen and hydrogen.
This battery was not of course very strong, nevertheless
decomposition of water could be obtained by it, and consequently
it presented an interesting voltaic circle, the synthesis and analysis
of water being exhibited at the same time, in one experiment:
the synthesis producing the voltaic current, shown by the gradual
absorption of the gas in the test tubes; the analysis produced by
the passage of the current in the exterior circuit, demonstrated
by the liberation of oxygen and hydrogen gas in a voltameter.

5. Voltaic Polarization has since been the object of interesting
labours on the part of a great number of scientists, notably
Faraday, Wheatstone, Schcenbein, Poggendorff, Buff, von Beetz,
Svanberg, Lenz and Saweljew, Edmund Becquerel, du Moncel,
Gaugain, &c.

The object of these works has been, in general, studying or
measuring the secondary current obtained with platinum electrodes.

We remember, however, that Mr. Sinsteden,(r) in a memoire
upon the results obtained with a magneto-electric apparatus,
having by haphasard sent the current from this apparatus through
voltameters with lead, silver, and nickel plates, obtained secondary-
currents with these metals, sufficiently intense to raise wires
to a state of incandescence.

(i) Sinsteden. Recherches sur le degre de force et de continuite du courant d'un
grand appareil magneto-electrique de rotation. Annals de Poggendorff. t. LXXXXII,
p. 16, 1854.

6. The study of secondary currents produced by

various voltameters.— The researches we have, on our part,
undertaken in this subject, in i859,(I)were made with the special
object of comparing the secondary currents produced by volt-
ameters of various metals in different solutions, by taking
observations directly after breaking the primary circuit, and at the
same time, studying the details of the phenomena manisfested in
these voltameters, with regard to the development of a secondary
current.

Figure 1 shows the voltameter or voltascope used. This
apparatus was arranged so as to enable the effects produced
around the electrodes to be easily followed (Fig. i). We took
care to light it well, and its transparency admitted of perfect
observation. It was arranged to take four wires, for studying,
in certain cases, the part taken by each electrode in producing

the secondary current.

All observations were taken
under the same conditions, by
employing wires of one diameter,
immersed in the same quantity
of liquid, placed an equal distance
apart, and subjected for the same
time to the action of the primary
current.

A mercurial commutator
(Fig. 2), similar to that of
Ampere, closes the secondary
1 circuit through SS', the instant
~ the primary curcuit PP' is broken.

Fig. i.

(i) Recherches sur la polarisation voltaique. Comptes-rendus, t. XLIX, p. 402,
1859. Bibliotheque universelle de Geneve, t. VII, p. 292, 1860.

Fresh researches upon polarisation were again made since this period, by Messrs,
du Bois-Reymond, Crova, Raoult, Thomson, Paalzow, Patry, Parnell, Branly, Thayer,
Lippmann, Bernstein, Fleming, Colley, Hankel, etc,

6

Fig. 4 shows the general arrangement of the apparatus.

P is the primary battery,
composed of from two to four
Bunsen cells, / is the commu-
tator, Fthe voltascope. Fig. 2.

At G, is a galvanometer, or tangent compass, allowing the
variations in intensity of the primary current to be followed, as
polarisation is developed in the voltameter.

The maximum polarisation was obtained rather rapidly, as the
wires only presented a very small surface, and the resistance of the
galvanometer was very 'low.

^ 3-

At G1, G2, G3, were placed three galvanometers of different
degrees of sensitiveness, for taking observations of the secondary
current. One of these galvanometers was replaced by a tangent
compass, when it was desired to measure the intensity, and when
the indications of the other two galvanometers had given an approx-
imate idea of the power of the secondary current. This current
only possessing, in certain cases, a very short duration, the deviation
was noted by placing beforehand a stop behind the needle, at a
point very close upon the maximum degree which it could attain,
which was determined approximately by several preliminary
experiments. This stop was then moved forward until the needle

could only make a slight movement of about half a degree further,
after closing the secondary circuit.

7. We have thus recognised that the order in which metals
used as electrodes in a voltameter, with water acidulated by
sulphuric acid, could be classed, with a regard to the intensity of
the secondary current, observed immediately after breaking the
primary current, was not exactly the reverse of the order in which
these same metals were classed with regard to the intensity of the
primary current passing through the voltameter, such as would
have been the case, if the tendency to the production of a
secondary current, or the polarisation E.M.F. as generally termed,
was the principal or only cause of the enfeebling of the primary
current.

Thus, to give the most striking example, we have noticed that
the voltameter with aluminium electrodes is the one which most
weakens the intensity of the primary current, for it ends by
almost completely arresting the passage of the current after a
certain time.(l) This metal would be consequently placed last on
the list of metals, classed in the order of intensity in the primary
current passing through the voltameter. If the cause of the
enfeebling of the primary current were only due to the secondary
E.M.F., aluminium should give the strongest secondary current,
and would consequently be placed first upon the list of metals
classed according to the intensity of the secondary current which
they give in the voltameter. Now, it is not thus, for aluminium
is found to give a weaker secondary current than any other metal.
This arises from the fact that other causes, often more influential
than the secondary current itself, such as the formation of a layer

(i^ Comptes-rendus, t. XLIX, p. 403, 1859.

UNIVERSITY

8

of oxide at the positive pole, and its insolubility and want of
conductivity, also the sheaths of liquid arising from the action of
the metal upon the electrolyte, at one or other of the poles,
contribute towards weakening the primary current.

This is the result of the careful examination we have made of
the phenomena which take place in voltameters composed of
various metals and different liquids. (2)

8. Voltameter with copper electrodes.— In a volt-
ameter, copper presents a seiies of phenomena which clearly
show the influence exercised by the oxidation of the metal at the
positive pole, and by the more or less dissolving action of the
liquid surrounding it upon the intensity of the primary current.

Fig. 4.

Fig. 6.

(2) M. Gaugain in a work upon the secondary E.M.F. of platinum plates in
voltameters (Comptes-rendus, t. XLI, p. 1165, 1855), showed the interest there was in
studying the effects of this force, after the electrolysis, that is to say, after the action
of the primary current. M. Gaugain concluded that the diminution in intensity
resulting from the presence of an electrolyte in a voltaic circuit was a phenomenon
much more complex than had been hitherto supposed,

If the current from one or two Bunsen cells be passed through
a voltameter with copper electrodes, in water acidulated by
sulphuric acid, three distinct phases are observed :

First, immediately the connections are made, the positive wire
becomes darkened without liberating any gas, the negative wire
freely liberates hydrogen gas, and the galvanometer makes a
strong deviation.

Secondly, the liberation of hydrogen decreases and is almost
stopped for a moment. The needle of the galvanometer returns
very nearly to zero. This phase corresponds to the formation of
the full thickness in the layer of oxide of copper round the
positive pole (Fig. 5.) ',

Thirdly, the current being stopped, the oxide of copper begins
to be dissolved; the positive wire is stripped of the greater part of
the layer of oxide which envelopes it, and gives rise to a flow of
sulphate of copper, in the form of a dense bluish liquid line,
towards the lower part of the cell. At the same time the liberation
of hydrogen recommences at the negative electrode; but it is less
abundant than during the first period (Fig. 6). The deviation of
the galvanometer likewise increases, but it is far from attaining the
original point.

The current having resumed a certain intensity, it appears that
a fresh layer of oxide ought to be deposited upon the electrode,
and then the phenomena of stopping and recommencing would
take place indefinitely, so long as the passage of the primary
current continued. But the layer of copper sulphate, formed upon
the wire, produces a movement in flowing away which prevents the
oxide adhering in a thick layer, and allows of its being dissolved
to a great extent, as quickly as it is formed. The positive wire
once stripped of the layer of oxide formed, apparently does not

10

become recoated, and the copper continues to be dissolved
without the preliminary phase of oxidation. The metal is only
slightly discoloured.

The intensity of the current during the third period, is reduced
to about one sixth of what it was during the first moments;
consequently equilibrium results between the two actions which
tend to be produced nearly simultaneously; namely, the formation
of a bad conducting oxide, under the never ceasing influence of
the current, and the solution of this oxide by the acidulated
water in the voltameter.

Nevertheless, this proportion between the original and final
intensity is not always maintained in the same degree; it varies
according to the duration of the experiment; because, the liquid
becoming gradually charged with copper sulphate, the negative
wire becomes coated with reduced, spongy, copper; the conducting
power of the voltameter increases in consequence of this deposit,
and the intensity of the primary current naturally tends to increase.

9. The secondary current, tried immediately after each of
these three phases, practically showed the same intensity; which
shows that the remarkable variations in intensity in the primary
current of which it has just been question, do not mainly depend
upon the secondary current.

This current is besides very weak in a voltameter with copper
electrodes and water acidulated with sulphuric acid. In order to
measure the E.M.F., the same as in voltameters of some other
metals, immediately after breaking the primary current, we have
used BecquerePs electro-magnetic balance, by ascertaining, in a
series of successive tests, the weight which it was necessary to
place beforehand in the scales, in order that the secondary current
might still produce a slight movement of the beam of the balance

11

during the first moment of passing. We have thus found that the
E.M.F. of the secondary current given by a voltameter with
copper electrodes, and water acidulated by sulphuric acid,
immediately after breaking the primary current, was only about
one tenth that of a Daniell element.

10. If, during the passage of the primary current through a
voltameter with copper electrodes and water acidulated with
sulphuric acid, when the intensity is reduced to that of the third
phase, the positive electrode be shaken, the liberation of hydrogen
becomes more abundant at the other electrode; the primary
current increases and keeps, to a certain extent, constant during
the whole time that the movement continues.

The effect of this disturbance is to separate the layer of saline
liquid which surrounds the wire and to favour the solution of
oxide of copper by the acidulated water, as it becomes formed.
This layer is then an important cause of the falling off in the
primary current, and has more influence in the case in question
than the tendency of the voltameter to produce a secondary
current; for this latter current, far from being weaker after the
disturbance of the voltameter, also increases in intensity. This
fact is owing to the sulphate of copper being rapidly dispersed in
the liquid of the voltameter during its disturbance; the negative
electrode becomes coated as described previously (8) with a spongy
deposit of reduced copper, and this wire, thus altered, is found to
be in a favorable state for the production of a more intense
secondary current, as we shall explain further on in detail (52),
when treating of the chemical actions in secondary cells with
lead plates.

As regards disturbance of the negative electrode in a copper
voltameter, it has no effect; for it only displaces the liquid a little

12

more (already set in motion by the bubbles of gas liberated round
the wire), and the liquid does not otherwise exercise any action by
itself upon the negative wire.

11. If, instead of water acidulated by sulphuric acid, we use
concentrated sulphuric acid in a copper voltameter, the layer of
oxide formed at the positive pole is not dissolved at all and
preserves the underlying metal from any attack, and it remains
unchanged under the influence of the current. Sulphate of
copper (CuO,SO3,5HO) cannot be formed, there not being the
necessary water in the concentrated sulphuric acid for its
formation. The current is gradually reduced to zero, in
consequence of the resistance of the oxide.

12. Voltameter with silver electrodes.— The phen-
omena observed in a copper voltameter are reproduced with a few
modifications in a voltameter with silver electrodes. We observe
the three phases, of maximum, minimum, and recommencement
of the current, especially when only employing a single Bunsen
cell as primary source.

The positive wire becomes a deep grey colour, or blackened ;
but the period of decrease only lasts a very short time, so that it
is more difficult to measure than with copper. The positive wire
is not completely stripped of the layer of oxide as quickly as it
is formed; it remains covered in spite of the flowing away of the
saline liquid.

The secondary current is much more intense than in the case
of copper; but it is also of short duration, in consequence of the
spontaneous solution in the liquid of the layer of oxide which
constitutes the main source of the secondary current, as will be
seen further on.

13

13. A phenomenon is shown with silver which is not perceived
in the case of copper. When the primary current is cut off, the
positive electrode which has preserved a thin deposit of oxide
upon its surface, gives rise to a liberation of gas, even when the
circuit is not closed. This liberation arises from local action
taking place between the oxidised surface of the wire, and the
metal underneath it. It is known that electro-chemical deposits,
especially of this nature, are easily penetrated. The liquid then
comes into contact with the metal and its oxide at the same time ;
the oxide soon disappears, and leaves visible the silver wire in a
metallic state. If, at that time, the secondary circuit is closed, no
current is noticed.

This fact proves that the secondary current must arise
particularly from the chemical action produced at the positive
electrode.

To confirm it, a third silver wire, not intended to have the
primary current passed through it, is immersed in the voltameter.
It is arranged so that at the moment the secondary circuit is
closed, it becomes connected with the oxidised positive wire,
whilst the negative wire remains out of the circuit. Now we
observe that the secondary current exhibits nearly the same
intensity as when the two electrodes are submitted to the action
of the current. The negative electrode does not then contribute
sensibly to the production of the secondary current.

14. The effects observed by causing a disturbance in a
voltameter with silver electrodes, are the same as those presented
in the copper voltameter. The dissolved silver salt is distributed
in the solution, the negative wire becomes covered with a spongy,
metallic deposit, and the secondary current becomes noticeably
increased.

14

15. In a silver voltameter with concentrated sulphuric acid,
the positive silver wire oxidises, and becomes gradually dissolved,
the reverse of what takes place with copper in the same liquid (i i).
This is explained by the chemical composition of sulphate of
silver (AgO,SO3) in which combination water does not take part.

16. Voltameter with tin electrodes. — The three phases

are very clearly marked in the case of this metal. The positive
wire blackens, and gives rise to an abundant flow of protoxide(t)
sulphate of tin. The secondary current, although less strong than
that from silver, has nevertheless considerable intensity. Disturb-
ance produces, as with the preceding metals, an increase of the
primary current, a deposit of tin upon the negative electrode, and
an increase in the secondary current.

17. The lead wire voltameter.— This voltameter does

not show the three successive phases in the intensity of the
primary current that is noticed with the preceding metals. This
is because the peroxide formed round the positive pole is
insoluable and coats the electrode in a permanent manner,
without gradually disappearing in the solution like the metals
already studied.

This adherence and insolubility of the peroxide of lead, added
to its affinity for hydrogen on account of the high degree of its
oxidation, contribute to produce, in a voltameter with lead
electrodes, a more intense secondary current and longer in
duration than that of any of the other metals.

18. Lead covered with peroxide of lead, in water acidulated
by sulphuric acid, acts in fact in a manner exactly the reverse to
that of zinc in the same liquid. It tends to decompose the water,
by absorbing hydrogen, and to become the positive pole of a cell,

Query by Translator. — Stannous Sulphate.

15

if it is connected with lead not oxydised, whilst pure zinc tends
to decompose the water, by absorbing oxygen, and becomes the
negative pole of a cell in which it is opposed to another metal.

To this cause of the development of a secondary current by
the voltameter of lead electrodes, we may add the effect produced
upon the wire or plate of the negative pole on short-circuiting
the voltameter after being submitted to the primary current.

The lead plate placed at the negative pole does not undergo,
by the action of the primary current, as marked a change as that
of the positive pole; nevertheless, as lead is always more or less
oxydised by exposure to the air, it is brought to a more perfect
metallic state by the hydrogen which is manifestly the means of
reducing the cell, and its tint changes from bluish grey to a much
lighter grey.

When we then close the secondary circuit, water being
decomposed in the cell, simultaneously with the appearance
of hydrogen upon the peroxidised plate, oxygen is carried to
the plate rendered formerly metallic by the primary current,
and oxidises it lightly. This oxidation is even visible; as the
negative lead plate immediately darkens upon the closing of
the secondary circuit. A single plate of lead,, or one united
with another plate exactly the same, would be thus oxidised in
water acidulated by sulphuric acid, and it would not give any
E.M.F., no more than would pure or amalgamated zinc. But
as the union of pure or amalgamated zinc with another metal less
easily attacked, both immersed in acidulated water, or better still,
in a liquid that can combine with hydrogen, occasions the zinc to
be attacked, and consequently the development of a current ; so
the union of an ordinary lead plate with a peroxidised lead plate,
which tends to decompose water by absorbing hydrogen, induces

16

at the same time, oxidation of the other plate, and consequently
the developement of an extra E.M.F. arising from this oxidation.

19. Such is the double chemical action which takes place in
a voltameter of lead electrodes, upon the closing of the
secondary circuit after breaking the circuit of the primary current.
And such is the double cause of the development of the strong
secondary current afforded by this metal.

If the circuit of the secondary cell be not closed after breaking
the primary circuit, there is produced nevertheless, a visible
chemical reaction in the voltameter, like that we have already
described in the case of silver (13). There is a slight liberation
of gas for some seconds at the positive plate. This fact is
explained, in like manner, by the affinity of peroxide of lead for
hydrogen, which causes local action to take place between the
oxidised surface and the metal beneath, and, consequently,
decomposition of the solution.

The action of metallic peroxides in absorbing hydrogen carried
from the electro negative element, and increasing the E.M.F. of
primary batteries, had been also noticed by Becquerel in regard to
peroxide of manganese, and De la Rive had obtained, by means of
powdered peroxide of lead, heaped around platinum or carbon,
a cell of a higher intensity than either Grove's or Bunsen's.(l)

20. We have measured, several times, the E.M.F. of a
voltameter with lead electrodes completely polarised by a
sufficiently prolonged action of the primary current. We have
found, by operating immediately after breaking the circuit of this

(i) Archives de 1'Electricite, t, III, p. 159, 1843.

17

current, that it was approximately equal to 1*5, a Bunsen element
being taken as a unit.(l)

This E.M.F. has since been measured by M. Edmond
Becquerel,(2) during the actual passage of the primary current, so
as to determine, by the difference, the fall caused in the E.M.F. of
the battery by the interposition of the voltameter with lead
electrodes. This E.M.F. was thus found equal to 1*41, the
current from the Bunsen cell being i.

21. The disturbance of one or other of the electrodes in a
voltameter of lead wires or plates does not produce the effects
observed with the preceding voltameters, which is easily explained
if we consider that the effects are due to the separation of the
liquid in different degrees of density, arising from the decompos-
ition of the various metals, and that the lead remains coated with
an insoluble oxide in the liquid in which it is immersed.

22. If we use salt water in the voltameter, instead of water
acidulated by sulphuric acid, there is formed round the negative
pole chloride of lead, scarcely soluble, and a very bad conductor,
so that the primary current is rapidly diminished, and the
secondary current itself is much weaker than that which is produced
in water acidulated by sulphuric acid. We have found that the
E.M.F. of this current measured, as in the preceding cases,
immediately after breaking the primary circuit, was but "08 that
of a Bunsen element.

This result explains why Ritter, who nearly always used salt
water for solution, did not notice any marked effect with lead, and
did not employ this metal in his secondary batteries.

(1) Comptes rendus, vol. 50, p. 640, March 1860.

(2) Annales du Conservatoire, No. 2, p. 277, October 1860,

18
23. Voltameter with aluminium wires. — Aluminium

shows, in a marked manner, the failing which takes place in the
primary current by reason of the insolubility and resistance of
oxide formed in the voltameter. (I) .^ >^^—

Two cells of this metal give a strong deviation; but it quickly
disappears, and there remains but a very feeble current.

It is even the same with four cells. It may be easily proved
that this diminution is owing to the bad conductivity and
insolubility of the oxide.

First, we show that, when the current has become very weak,
if the circuit be broken for a few moments, and then closed again,
the current does not regain its original intensity.

That being so, if we replace the positive wire by a freshly
brightened' or new wire, a very decided deviation will be observed,
which disappears in its turn.

Changing the negative wire only causes a hardly noticeable
increase of current.

22. The positive wire, taken out and examined immediately
upon the decrease of current, shows no change in color or
appearance, although it is probably oxidised. If it be washed,
thoroughly wiped and replaced in the voltameter, it does not allow
the current to pass any more easily; it must be rubbed with glass
paper before a momentary but decided deviation is observed.

Shaking has no effect upon either wire.

The secondary current is extremely weak with this metal, so
much so, that the great decrease in the primary current cannot be
attributed to the opposing E.M.F. of the secondary current.

(i) Comptes rendus, vol. XLIX, p. 403. (1859), and Bibl-univ; de Geneve, vol. VII,
p. 292, 1860.

19
24. Voltameter of iron and zinc wires. — We have not

spoken until now of the primary current which may be given by
the voltameter, because the preceding metals are not sensibly
attacked in acidulated water. Yet, when both wires are very well
brightened, they sometimes give a slight deviation with a sensitive
galvanometer, but this effect is not worthy of notice compared
with those we have taken into consideration.

With iron and zinc, which decompose rapidly in acidulated
water, the least difference in the attack of the two wires gives a
current strong enough to be worthy of attention; besides, the
direction of the current changes every minute, according to the
degree in which each wire is attacked. We cannot then easily
determine either the quantity or the E.M.F. of the secondary
currents which are produced. There is however a means of
observation which we may use, in order that the effects may be as
little disturbed as possible by the primary current of the
voltameter; it is to abstain from dipping the electrodes into the
liquid beforehand, and to arrange the connections so that the
circuit may be closed by the act of dipping the electrodes into the
acidulated water. By operating in this manner, the electrodes
become immediately submitted to the action of the primary
current, without having undergone any preliminary attack by the
solution itself.

We then find, more especially with iron, the three phases of
intensity in the primary current, which the greater number of
metals present. The positive electrode is darkened, and produces
an issue of sulphate of iron, or of zinc, without liberating any gas.
But upon breaking the circuit, the solution acts freely upon the
metal, which immediately discharges hydrogen, and with so much
the more energy as its surface remains slightly oxidised, The

20

negative electrode on the contrary, having attained a more
metallic state, under the influence of the electrolytic hydrogen,
produced by the primary current, remains some time in the body
of the acidulated water, without liberating any gas; but little by
little it necessarily ends by becoming attacked. We may consider
that the electrolytic hydrogen, by enveloping the metal, and
preventing its oxidation by the solution during the passage of the
primary current, plays the same part as mercury does in
amalgamated zinc.

The secondary current, tried by working as we have described,
is rather strong, and may be taken, during the first few moments,
as almost entirely due to the action of the primary current, but
the attack upon the electrodes by the solution itself soon disturbs
the effects.

25. Voltameter with gold electrodes. — The positive

gold electrode becomes visibly oxidised under the action of a
rather weak primary current (2 Bunsen elements), and becomes
coated with a reddish deposit of oxide of gold. If we change the
direction of the current, the oxide is rapidly reduced by the
hydrogen; the electrode becomes blackened without showing the
metallic lustre. Another change in the direction of the current
causes a fresh oxidation, still more rapid, on account of the finely
divided state of the metal, but it is no longer possible to tell by
its appearance, whether it is oxidised or reduced.

The secondary current is rather strong, but not very susceptible
of an exact measure, because of its short duration. Unlike what
happens with the other metals, it appears due, to a great extent,
in spite of the decided oxidation of the positive electrode, to the
action of hydrogen on the negative electrode. We recognise it in

21

working as described previously (13), that is to say, by placing
beforehand in the voltameter, a third and a fourth electrode
outside the principal circuit. These electrodes being successively
coupled with the positive and negative electrodes, immediately
after the action of the primary current, we find that the secondary
current given by the negative electrode and one of these
nonpolarised electrodes, is stronger than the current produced by
the positive and the other nonpolarised electrode. We explain
this fact further on (27, 28, 29,) when treating of the platinum
voltameter giving the same result.

26. Platinum Wire Voltameter. — The phenomena
produced by the previously studied voltameters may help to
account for those which are produced by the platinum wire
voltameter, the explanation of which is also not devoid of
difficulties.

If the primary current is supplied by a single Grove or Bunsen
cell we know that electrolysis cannot take place, and that the
current is almost completely reduced to zero by putting the
voltameter in the circuit. If the primary current is supplied by
two cells, electrolysis takes place, and the current, after shewing
some decrease, is maintained constant without first going nearly to
zero, as happens with metals easily oxodised. Platinum does not
visibly oxidise at the positive pole, although the secondary current
produced is strong ; it is decidedly stronger than that given by the
gold wire voltameter, but like the latter, it is of too short a duration
for anyone to take an exact measure of the E. M. F. directly
after the breaking of the primary circuit.

27. It proves, the same as with gold, that the action of the wire
which evolves hydrogen during the electrolysis helps much more

'UNIVERSITY;

22

to produce the secondary current than that of the wire which
evolves oxygen.

28. This fact is explained by the condensation — or even
combination of the hydrogen gas with the platinum, — which seems
to behave like a metallic vapour, as Prof. Dewar has drawn atten-
tion to long ago. This combination being of course extremely
transient and easily oxidised, we see that it may become the source
of an electric current, and form the negative pole of a battery
element, in relation with another wire of the same metal which has
not undergone a similar action.(I)

Again the action of the electrolytic gases does not act only
upon the metal electrodes but also upon the solutions in the
immediate neighbourhood of the electrodes, producing some very
unstable combinations with the liquid itself. The oxygen liberated
round the positive platinum electrode in water acidulated with
sulphuric acid forms ozone from the oxygenised water, and as
recently shown by M. Berthelot,(2) a new compound, per-sulphuric
acid, the discovery of which we owe to him.

These various bodies, all highly charged with oxygen, must
contribute towards the partial production of the secondary current
by the positive electrode.

29. The hydrogen liberated around the negative electrode
forms on the other hand a compound corresponding to the
oxygenised water, that is to say, a body more highly charged with

(i) Graham's Works have shewn, as is known, this tendency of Hydrogen to ally itself
with metals, especially in the case of palladium ; Prof. Dewar has observed a strong secondary
current with a plate of hydrogenised palladium in connection with an ordinary plate of the
same metal, and the researches of Messrs. Crova, Root, &c., have proved that a similar
effect was possible with platinum.

(a) Comptes, Rendus, Vol. 86, pp. 20, 75, 1878.

23

hydrogen than the water itself, such as a sub oxide or protoxide of
hydrogen, which would further contribute towards the production
of the secondary current. Although a body of this kind has not
yet been set free, the strongly reducing or oxidising tendencies of
the electrolytic gases and the chemical symmetry of the phenomena
which take place at the two poles of the cell, permit of the idea
that there is a partial combination of the gas with the liquid round
the negative electrode, and in consequence the formation of a
compound of this nature.

30. There may also be another cause for the part taken by the
positive pole in the developement of the secondary current by
platinum electrodes, which there is occasion to mention : it is the
oxidation to a very feeble extent of the platinum itself by the
oxygen of the voltameter.

This oxidation, it is true, is not visible, but as it is so very
evidently produced on all metals — even gold itself — by the active
oxygen from the electrolysis, we are naturally led to think that
platinum cannot be completely free from this strong oxidising effect.

It was the opinion of De La Rive, who, after having first
admitted a simple physical action of the gases upon the electrodes,
afterwards thought that there must be chemical action produced
upon the platinum.

He had even observed the deterioration of electrodes of this
metal which had served for the passage of a current in a
voltameter alternately in both directions, and had rightly concluded
that platinum, in order to arrive at this state, must have undergone
successive periods of oxidation and reduction.

The fact given previously (23), as to aluminium not changing in
appearance at the positive pole, although it oxidises to such an

24

extent as to entirely stop the passage of the current, proves,
moreover, that a metal may be coated with a very thin layer of
oxide without altering its appearance.

A further proof in support of the fact of a very thin coat of
oxide being formed upon the positive pole of the platinum
voltameter is shewn by a platinum positive plate, washed and dried,
without, of course, being rubbed hard, preserving the power of
giving a secondary current when coupled with an ordinary plate,
the same as the positive electrode of ah aluminium voltameter,
likewise washed and dried, retains the property of very nearly
stopping the passage of the current, on account of the very thin,
invisible coat of oxide with which it is covered.

Thus, the chemical change in platinum, produced even to a very
feeble extent by the primary current, can assist, in some degree,
towards the developement of a secondary current.

31. We may finally add to the above causes the action upon
the electrodes of the gases in the solution which play an important
part, particularly in Groves' gas battery, as is also shown in the
works of Messrs, von Beetz, Gaugain and Morley.

Such is the total of the various causes which help to produce a
secondary current in the platinum voltameter, in particular, and
which apply equally to that with gold electrodes. They may
possibly apply also to a slight extent, with regard to other metals;
but we have seen that the oxidation of the positive electrode
played the principal part in the voltameters previously considered.

32. Voltameters with acidulated water, saturated

With bichromate Of potash. — By using, instead of water
acidulated with one tenth part sulphuric acid, a saturated solution of

25

bichromate of potash, equally acidulated, we notice phenomena
which further demonstrate the influence of insoluble layers of
matter deposited round the electrodes, or that of liquid sheaths,
in increasing the resistance and diminishing the primary current
in voltameters.

In this solution, silver and mercury become covered with deposits
of insoluble red chromates which very nearly stop the passage of
the current.

With the other metals, the reduction of the solution by hydrogen
liberated around the negative pole, tends, no doubt, to increase
the E.M.F. of the primary current; but, on the other hand, the
reduced liquid forms, round this electrode, a sheath lower in
conductivity, contributing to diminish it.

The secondary E.M.F. of these voltameters is a less important
cause of falling off in the primary current than is the change in
the solution. When the oxide of the metal is soluble, there is
produced at the positive pole another saline sheath, equally liquid,
which is only gradually dispersed and forms a further hindrance to
the passage of the current.

The movement of either electrode also produces a very marked
effect in these voltameters ; the decomposed layers of solution are
dispersed, the active solution finds its way more quickly to the
electrodes and the current regains nearly all its original intensity.

This is, besides, what is noticed in Poggendorffs bichromate of
potash battery, to which a fairly constant E.M.F. is given by the
injection of air into the solution ; — an improvement made by
Messrs. Grenet and de Fonvielle.

By studying the phenomena which take place in voltameters we
are able to account for those which are produced inside voltaic

26

cells; because they are but the model, transposed, so to speak, into
another apparatus where they may be examined under more
favorable conditions.

33. Conclusions. — We may draw, from the above, the
following conclusions :

The falling off in an electric current when a voltameter of
acidulated water and electrodes of various metals is placed in the
circuit, is due to several causes, acting with more or less intensity,
according to the metals used. Sometimes these causes are all
united.

1. The insolubility and bad conductibility of the oxide
formed upon the positive pole.

2. If the oxide be soluble, a bed of saline liquid is produced
by the decomposition and forms a sheath which prevents the
the rest of the solution finding its way to the electrode.

3. The inverse, secondary, current (polarisation) which arises.

34. This secondary current, observed on closing the voltameter
circuit, immediately the primary current is cut off, arises also from
several causes :

1. With most metals, it is due, principally to the reduction
of the oxide coating formed upon the positive electrode by the
action of the primary current, and to the oxidation of the negative
electrode which is brought to, or maintained in, a perfect metallic
state by the liberation of hydrogen under the influence of the
primary current.

2. With metals not easily oxidised, such as gold or platinum,
the secondary current is due, for the most part, to the action which
hydrogen has upon the negative pole, during the passage of the

27

primary current, whether by uniting itself with the metal of the
electrode, or by modifying the chemical composition of the liquid
with which it is surrounded, or by simply becoming dissolved in it
to a small extent.

The compound thus formed, the solution thus changed, or the
hydrogen gas dissolved, tend to combine again with the oxygen
arising from decomposition of the water whilst the circuit of the
voltameter was closed, and consequently furnish one of the elements
of the E.M.F. of the secondary current.

But this secondary current is also due, at the same time, although
in a less degree, to the slow oxidation in the positive electrode of
the metals in question, during the passage of the primary current,
on the one hand ; and on the other, to the formation of strongly
oxygenised compounds with the liquid in the voltameter; and
finally, to a small proportion of oxygen gas dissolved in the water
round the electrode. The slightly oxidised metal, the modified
solution and the dissolved oxygen re-combine with the hydrogen
when the secondary circuit is closed and furnish another element
of the total E.M.F. of the voltameter.

35. The causes which contribute towards the developement of
the secondary current in a voltameter are, as may be seen, very
numerous. We could, strictly speaking, mention many more of a
purely physical nature, resulting from the electrical condensation
which must always take place in a system made up of two conduct-
ing materials of a certain nature, separated by one of a different
nature. But, although the idea of a simple physical action may have
arisen in the minds of physicists, in order to explain the production
of a current without any apparent chemical action, and which may
have given rise to the expression voltaic polarisation, this kind of

28

action may be entirely disregarded here, as much on account of the
conductibility of the solution as the proximity of the surfaces in
play, and the chemical actions just reviewed appear to us to be the
principal causes of the production of the secondary current.

CHAPTER II.

Storage of the energy of the Voltaic
Battery by means of Secondary Cells
with lead plates.

Various arrangements of these cells. — Their electro-
chemical FORMATION or preparation. — Their
effects. — Preservation of their charge. —
Residues. — Returns.

36. Since we have known that secondary currents constituted
an important cause of the falling off in the E.M.F. of voltaic
batteries, attention was devoted towards the prevention of these
currents in the batteries themselves, and they were very fortunately
neutralised in the first bi-liquid, and constant current batteries, due,
as we know, to Becquerel.(l)

(i) Annales de Chimie et de Physique, ze serie, t. XLI, p. 24, 1829.

30

Taking it from another point of view, we have sought to collect
the secondary currents and to make use of them, thus accumulating
the energy of the voltaic battery.

We have found, as may be seen in the preceding researches (20),
that the secondary E.M.F. of a voltameter with lead plates in water
acidulated with sulphuric acid, was higher and more lasting than
that of any of the other metals, and that it surpassed that of the
strongest voltaic element known — that of Grove or Bunsen.

With such an E.M.F., it only remained, in order to form a
secondary element of great power, to endow it with a very low
resistance, or to increase its surface to the greatest possible extent.
This was so much the more easy from
the fact that the two plates necessary to
form it were of the same nature and of a
metal so extremely flexible and malleable
as lead.

37. Secondary cell with coiled

lead plates. — It is thus that we were led
to construct, in i86o,<I)a secondary element
of great power or quantity, -by using an
arrangement similar to that which Offer-
shaus and Hare had employed for the
voltaic cell proper, namely, by rolling two
long, wide lead plates into a coil,
separated one from the other by a thick
cloth, and then immersing them in a
glass jar full of water acidulated with a tenth part sulphuric acid.

Figure 7 shows the construction of a secondary cell of this kind.

(i) Comptes rendus, t. L, p. 640. Mars 1860.

31

Fig. 8.

38. Secondary battery of large surface for quantity.

Figure 8 represents a secondary battery of nine cells,
having a total surface of ten square metres, the earliest
effects of which we exhibited at the Academic des Sciences, on
the occasion of the meeting of March 26th, 1860.

By passing through this apparatus the current from five small
Bunsen cells, we obtained, after a few minutes' action, a very bright
spark possessed of great heating power when the two terminal
wires of the battery in which the cells were coupled up, either in
three parallels of three cells each in series, or all in parallel as
shown in figure 8, were brought into contact for an instant.

39. Secondary cells with parallel lead plates.— But

the arrangement of secondary cells in the form just described

32

was somewhat objectionable by reason of the additional resistance
introduced by the cloth intended to separate the electrodes.
Moreover, this cloth becoming rotton in time, in the acidulated
water, the lead plates might come into contact with each other,
and so throw the cells out of order.

To overcome these objections, we used another arrangement'1'
consisting of two series of parallel lead plates, with the terminals
of the even row of plates joined on one side, and those of the
odd row joined together on the other side, put into communica-
tion with the two poles of a primary battery.

These plates, brought very near to each other, and separated in
the middle by insulating rods, were arranged vertically in a rectang-
ular gutta percha cell, furnished with interior grooves, for holding
the parallel lead plates.

In figure 9, which represents the plan of this arrangement, the
letters a b c, and a' b' c' show the two series of lead plates.

Figure 10 represents the apparatus complete. The
terminals of the plates joined together at K and K' come
through the upper part of the gutta percha cell, and may
be connected with the primary battery by the wires
P and P" . The cell having been filled with water
acidulated with a tenth part of sulphuric acid, a cover is
fused to the upper edges. A very little hole only, is
left, in order to allow an escape for gases arising from
Fig. 9. electrolysis, during the passage of the primary current.

(i) Comptes rendus, t. LXVI, p. 1255. Annales de Chimie et de Physique,

4th. series, t. XV, p. 10, 1868.

33

40. To illustrate the heating effects that can be produced with

this apparatus, a thick platinum or
iron wire / is fixed between the
terminals B and O, of which B is
connected with the plates a b <:, by a
small metal plate placed against one
of the sides of the cell, not seen in
the figure, and the other only put in
contact with the second series of
plates a b ' c through the intervention
of a contact maker M. This contact
^ maker is not intended to change the

Fig. 10. direction of the current, but simply

to connect up the circuit, sometimes with the primary battery for
charging, sometimes with the terminals O and B for discharging. (I)

41. By using six lead plates, twenty centimetres long by
twenty two high, and considering that the double surfaces of all
the plates are used, with the exception of the two outside surfaces,
we have thus a small secondary battery for quantity, with a surface
of about half a square metre, which will redden iron, steel, or
platinum wires one millimetre in diameter, after being submitted
for a short time to the action of a primary battery of two Bunsen
cells.

42. The above arrangement, which we have used for several
years, for obtaining at a moment's notice temporary currents of

(i) This arrangement of contact maker in the form of a simple metal bolt, rounded
at the ends, which fits into small cast brass tubes, seemed a very convenient method,
and we have used it since 1868, in several other ways, for making or breaking
electrical communication at will.

34

considerable intensity, appeared to still possess some objections,
which we have had to do our best to remedy.

The vertical gutta percha cells or jars became in time
contracted, and caused the lead plates to approach each other
very closely by buckling them, and causing contact.

This substance not being transparent prevented, besides, the
sight of phenomena which take place inside the cells, and which
it is important to be able to follow during the charge, as will be
seen further on (60).

43. Latest form of secondary cells with lead

plates. — We then returned to an arrangement very similar to the
first described (37), but modifying the method of separating the
lead,plates.(l) We separated these plates no longer with a thick
cloth, but by narrow strips of india rubber, presenting the
advantage of not being injured in the acidulated water and only
covering a very small portion of the surface of the electrodes.

Fig. ii.
Figure 1 1 represents the manner in which we proceeded to coil

(i) Comptes rendus, t. LXXIV, p. 592, 1872.

35

the plates close to each other without touching, and shows the
arrangement of a couple thus constructed.'1'

Two pairs of india rubber strips about one centimetre wide by
half a centimetre thick, are necessary to prevent the plates
touching each other. The terminals are shaped at the opposite
ends of the plates, in order to avoid as much as possible any
contact, and to equalise the distribution of the primary current
upon the surfaces of the electrodes, by separating the two points
by which the positive and negative electricity flow into the
secondary cell, as far as possible. But this is not indispensable if
the plates are very uniformly rolled together. The chemical
action of the primary current is then distributed equally over the
whole surface of the secondary couple, even when the two poles
are very close to each other.

We then coil the lead plates, thus separated by two or three
pairs of india rubber strips, round a wood or metal cylinder, as
shown in figure n. As this cylinder is withdrawn as soon as the
coil is finished, it does not matter of what it is made.

It is also better to place two small rubber strips at right angles
to the others along the cylinder when beginning to roll the first
turn, so as to well separate the edges of the two plates which
might otherwise come into contact.

44. The coil once made, we take out the interior cylinder with
care, and to make the whole more firm, the coils are permanently
held in position by means of little gutta percha crosses softened
by heat.

(i) We here enter into details, which will perhaps appear trifling, upon constructing
and putting these cells to work ; but the results achieved by their means in our more
recent researches, have led us to wish scientists to make these cells easily for themselves,
for the objects they might have in view. We are, besides, only acceding to a desire,

often expressed, for the knowledge of all the details of the secondary cells.

The couple, thus constructed, is put into a cylindrical
glass jar, and supported inside this jar by little gutta percha
wedges. The jar is then filled with water acidulated with one
tenth part sulphuric acid.

Fit. 12

45. — Figure 12 represents a secondary cell of considerable
size,(r) made according to our description, and also shows the
arrangement of the connections for charging or discharging the
cell, and illustrating the effects it can produce. (2)

(1) The lead plates are about sixty centimetres long, by twenty wide, and one
millimetre thick.

(2) Les Mondes, t. XXVII, p. 426, and following, 1872.

37

The glass jar containing the lead plates immersed in acidulated
water, is covered with a vulcanite disc which carries the metai
parts intended to close the secondary circuit, when the cell is
charged. The terminals of the two lead plates are connected by
means of the binding screws G and H, at the same time, with a
primary battery of two small Bunsen elements and with the little
copper plates MM'.(t) The small plate M is fixed under another
copper plate jR, the extension of which, forming a spring, can be
screwed down against the little plate J/, which is then connected
with the terminal A. The little plate M' on the other hand is in
fixed communication with the terminal A' , and between the arms
of these two terminals are placed the wires intended to be
reddened or fused by the secondary current. Wires from any
other apparatus through which it is desired to pass the same
current, may also be connected to these two terminals.

46. This system of connections, arranged in the simplest
possible manner, so as to be held upon the cover of a jar, does
not break the primary current when the secondary circuit is
closed, like that which has been described above (40). If then
we wish to try the effect of the secondary cell alone, it must be
cut off from the primary battery by disconnecting the wires
G and H, and then screwing up the button B. But if these wires
are left in communication with the secondary cell when it is
desired to discharge it by depressing the button, there is no

(i) The terminal tongues of the lead plates were previously covered whilst hot,
throughout their length, with a thick varnish of spirits of turpentine, or with a comp-
osition of wax and resin, in order to prevent the acidulated liquid from creeping by
capillary attraction, up to the connections with the copper plates. The terminals of
the lead plates at this point, must be thoroughly brightened, and then covered with
varnish or buried in a bed of composition poured upon a cork which seals the cell. This
cork is pierced with a hole through which a little glass tube is passed.

38

objection; the apparatus works just as well; its effects are even
slightly increased by the action of the primary battery, which is
added to that of the secondary cell, when the circuit of the wire
F is closed ; for the two currents from the primary battery and the
secondary cell, which are in opposition during the charge, become
joined in parallel in the discharge.

47. We shall see further on (75), that the secondary cell,
formed or prepared in a special manner, preserves its charge long

enough to give strong effects, without the additional help from the
current of the primary battery, during the short period of
the discharge. But the continued connection of the secondary
cell with the primary battery, presents, at least, the advantage that
if we have a series of experiments to make with successive
discharges of the secondary cell, all the intervals in which the
discharge is not going on, become immediately utilised by the
primary battery for charging the secondary cell.

48. Chemical actions produced in secondary cells

Of lead plates. — We have already studied (18) the chemical
actions which take place in a voltameter with lead wires or plates,
when the circuit of the voltameter is closed (short circuited) after
the action of the primary current, and we have considered in
detail the principal causes of the strong secondary current
produced.

In the secondary cells which we have just described, these
actions are naturally exhibited upon a larger scale, and produce
phenomena, the study of which has been useful for endowing
these cells with important qualities, such as furnishing discharges
of long duration, of preserving their charge for a long time after

39

the action of the primary battery, and of thus storing the
chemical work of the voltaic pile.

49. When a secondary cell of large surface, such as that shown
in Fig. 12, is new, that is, when the lead plates comprising it have
never served to transmit the current in a voltameter, and it
happens to have the current from two Bunsen cells passed through
it, oxygen gas appears almost immediately upon the positive plate;
some of it at the same time oxydises the surface of the plate, and
this becomes quickly covered with a very thin coating of peroxide
of lead.

On the other hand, the hydrogen, after having reduced the
slight layer of oxide with which the lead was probably covered by
exposure to the air, soon appears, and if, in a few moments, we
try the secondary current given by the apparatus, we find it is
already very strong by the sharpness of the spark produced when
the secondary circuit is closed and opened again instantaneously,
with a copper wire of low resistance. But the current thus obtained
is of very short duration. It would certainly produce the incandes-
cence of a very fine platinum wire, but would not redden a thick
wire of the same metal. This happens from the layer of peroxide
of lead on the surface of the positive plate being very thin, and as
it becomes quickly reduced immediately the secondary circuit is
closed, it cannot furnish a sufficient quantity of electricity.

50. But if, after having kept the circuit closed until the
secondary current is exhausted, the apparatus be charged a second
time, the plates are then in a condition little differing from that in
which they were at the commencement. Whilst the secondary
circuit is closed, the oxygen, being carried against the plate which

40

was negative during the passage of the primary current, has slightly
peroxidised this plate, at the same time that the peroxide formed
upon the other plate during the passage of the primary current,
was reduced by the hydrogen. We have then, after a first
experiment, two plates, the surfaces of which present a molecular
condition differing from that in which they were when new. They
are covered with thin layers of oxide and reduced metal
respectively, which will facilitate the ulterior action of the primary
current upon the secondary cell.

51. If we first consider the lead plate which was negative
when the current passed through for the first time, it is as we have
just seen, covered with a layer of oxide after the passage of the
secondary current. The consequence is that if the primary
current be passed through it again, the first portion of hydrogen
will be devoted to reducing this layer of oxide, instead of the
thinner layer resulting only from exposure to the air, as had
happened previously. Then a period longer than the first will
take place before the appearance of hydrogen upon the surface of
this plate; for this gas will not begin to be set free till the oxide is
completely reduced to a state of pulverulent or highly divided
lead upon the surface of this plate.

52. If we study what takes place upon the positive plate for
the second time, during the action of the primary current, the
first portion of oxygen which tends to be set free at its surface
encounters, this time, a layer of reduced peroxide or divided
metallic lead, upon which this gas has more hold, so to speak,
than upon the lead plate which serves for the first time ; the gas is
more easily absorbed, and we also begin to note a delay in the
appearance of oxygen upon this plate, a period which corresponds

41

to the time necessary for re-oxidising the layer of reduced lead
upon its surface.

When the secondary circuit is again made, the above described
phenomena again take place, and it is easy to imagine that when
these operations have been renewed a great number of times, the
surfaces of lead in the secondary cell will be found in a more
favorable condition for oxidation or reduction; the layers of
oxide alternately formed or reduced, will become thicker, and the
resulting secondary effects will show a longer duration and greater
intensity. This is in fact what is observed: the more a
secondary cell is charged and discharged, the greater is the
duration of the secondary current.

53. Formation, OP electro-chemical preparation, of

secondary cells Of lead plates. — We have thus attained the
extention in duration of the discharge of secondary cells, by
charging them successively a great number of times, and discharg-
ing them in proportion, so as to develope upon their surface, and
even produce to a certain depth in the thickness of the plates,
these layers of oxide and reduced metal, the finely divided state
of which favors the development of the secondary current.

This result has also been obtained in a still more marked
manner, by successively changing, several times, the direction of
the primary current acting upon the secondary couple.

Everytime this change of direction is made, we notice, if there
is a galvanometer in the circuit, a remarkable reinforcement of the
primary current, for the first moments, which is easily explained
because the E.M.F. of the secondary cell is added under these
conditions, to that of the battery. But it is better, in order not to

42

exhaust the primary battery too quickly, to short circuit the
secondary cell before reversing the primary current; without that,
the current from the battery is devoted for the first few moments
to doing chemical work which could be effected by the discharge
of the secondary cell itself.

Thus, the secondary cell must be first discharged, then recharged
in the opposite direction. In this case, the primary current again
acts directly upon surfaces which are nearly in the same electro-
chemical condition; we do not then observe the increase of the
current mentioned above, and there is no useless loss in the
energy of the primary battery in the form of heat dispersed
through the circuit.

54. We have seen moreover that it was advantageous, with a
view to this preparation of the secondary cells, to allow a period
of repose of several days between the reversals, in order to give
to the deposits of oxide and reduced metal, time to attain a
crystalline nature, and strong adherence to the surface of the
plates. We observe in fact that the plates of the secondary
couples which have undergone these actions acquire in time a
peculiar crystalline appearance. The peroxidised plate, and the
plate covered with reduced lead, are both of them strewn with
shining points, and become changed in their molecular construction,
not only at the surface, but some distance within the pores of the
metal; we even notice that the peroxidised plate in particular ends
by showing considerable fragility.

55. The plates thus changed lose none of their weight, by any
number of charges and discharges. They only act, so to speak,
as a base for the chemical actions which take place upon their
surface, and which, constantly following each other in opposite

43

directions, cannot consume them. The lead is continually
oxidised and reduced at the same time that the water is alternately
decomposed and vice versa.

56. These combined operations that we have called by the
name of formation of secondary couples, and which consist, as
we have just seen, in forming or maturing them, in order to
produce deposits of a certain thickness, allow of our obtaining
heating effects of considerable duration in the discharge.

Taking a secondary cell with lead plates of half a square metre
in surface, formed or matured to a desired extent, and previously
charged by two Bunsen elements for three quarters of an hour,
we can redden for eight or ten minutes a platinum wire one
millimetre in diameter, and seven or eight centimetres long, and a
wire of about half a millimetre in diameter, for twenty to twenty
five minutes.

57. Prolonged immersion of the lead plates in the acidulated
water, before the action of the primary current, hastens materially
the formation of secondary cells.

This fact seems to be explained by the slow penetration of
the liquid into the interior of the pores of the metal, which allows
the electrolytic action to be exercised deeper, and to produce a
greater quantity of oxide or reduced metal.

58. The intensity of the primary current also has a great
influence over the more or less perfect formation of secondary cells.
The current produced by two Bunsen elements is what experience
has proved to be the most suitable. A too feeble current only
produces very superficial deposits, and the nature of the peroxide

44

of lead produced upon the positive plate is different as to its physical
aspect (and perhaps in a chemical point of view),(l) from that of
the peroxide formed by a stronger current. The resulting oxide
from a sufficiently prolonged feeble current is black ; that which a
stronger current creates has a clear brown color characterising the
true peroxide of lead.

Daniell elements, even in a large number, do not form secondary
cells so well as two Grove or Bunsen elements possessing less total
E.M.F., but giving a larger quantity of electricity. The oxidations
and reductions must be made at a certain rate, and the electrolysis
must be sufficiently vigorous to thoroughly penetrate the interior
of the metal.

59. To form the secondary cells well, it is expedient to bear
in mind the preceding remarks, and to take into consideration
above all that the action of time is indispensable, as in many
chemical actions of which nature and industry present examples.

It is a kind of electro-chemical tanning, if one may use such an
expression, which the electrodes of secondary batteries undergo.
The lead plates should be penetrated little by little, as deeply as
possible by the oxidising and reducing actions of the primary
current, so as to completely change their molecular formation.

(i) We have had occasion to notice in voltameters with copper electrodes, an example of
the difference in the chemical nature of the oxide formed, according to the E.M.F. of the
current used. Thus with a current from two Bunsen elements, there is obtained at the
positive pole, black bioxide of copper, which only appears for an instant at the commence-
ment of the electrolysis (8) and is then dissolved in the liquid without being visible, according
to the degree of its formation ; whilst if fifteen elements are used a spray of reddish oxide
resembling protoxide of copper is formed at the extremity of the positive electrode, and
falls to the bottom of the voltameter without being immediately dissolved in the liquid.

(Bibl. univ. de Geneve, t. VII p. 332, 1860.)

45

The periods of rest mentioned above (54) between the changes
in direction of the primary current have the greatest effect. Thus
a secondary cell the plates of which have been submitted for
several hours at once to the action of the primary current, being
left to itself for a month, without being discharged, then taken in
hand again and charged in the reverse direction, will give a
discharge of double the duration which it could give before. The
following, in fact, is the plan to be pursued in these operations ;

The secondary cell being filled, to begin with, with water
acidulated by a tenth part of pure sulphuric acid,(l) we allow the
current from two Bunsen elements to go through it six or eight
times, in different directions alternately, on the first day. The
secondary cell is discharged between each change of direction'
and it is easy to take note, either by the incandesence of a
platinum wire or by any other plan, that the length of discharge
continues regularly to increase.

Gradually we increase the time during which the secondary cell
is submitted to the action of the primary current in the same
direction. This period is successively extended, after the first day,
from a quarter of an hour, to half an hour and an hour. We
leave it finally charged in one direction until the next day. The
following day it is charged for two hours in the reverse direction,
then in the first direction, and so on. We still notice an increase
in the time of the discharge, but a point soon arrives beyond
which this duration does not perceptibly increase, especially when
the primary battery, not having been renewed, becomes gradually

(i) It is essential that the sulphuric acid contain no traces of nitric acid, which would
attack the lead and help to cause the terminal tongues of the plates of secondary batteries
to be broken off.

46

weakened by these successive operations, and has no longer
sufficient intensity for the electrolysis to penetrate deeper into the
interior of the plates (58).

We then leave the secondary cell at rest for eight days, and it is
then recharged in the opposite direction for several hours, without
making for that day any fresh reversal. Gradually the period of
rest is extended to a fortnight, a month, two months, &c., and the
duration of the discharge continues to increase. There is no limit
to it but the thickness of the lead plates.

The positive plate, if it is thin, ends by being almost completely
transformed, in time, into peroxide of lead of a crystalline nature.
The negative plate becomes gradually changed, to a certain depth
beneath its surface, to reduced lead, spongy and crystalline.

It is not, of course, necessary to push the electro-chemical
preparation of the secondary cells to this complete transformation
of the physical and chemical nature of the plates ; for the plates
would then acquire a higher resistance, and require more time for
charging.

When the secondary cells give a current of sufficient duration
for the purposes to which they are to be applied, there is no further
occasion to change the direction of the primary current each time
they are used. The store of peroxide of lead accumulated on the
positive plate would take too long to reduce, and no effect would
be obtained from the cell for several hours. One direction is then
adopted in which the secondary cells are always charged, once they
are sufficiently formed.

60. Absorption of the gases during1 the charge of

Secondary cells. — When secondary cells prepared under the

47

foregoing conditions are charged, we observe that the gases are
completely absorbed, for a certain time, to such a point, that, with
a secondary couple having a surface of one square metre,
submitted to the action of two Bunsen elements, it takes twenty to
thirty minutes before any gas appears upon the surface of the plates.

The entire work of the primary current is stored in the apparatus
in the form of oxidation of lead on the one hand, and reduction
of the lead oxidised by the previous discharge of the secondary
current, on the other hand, and may be again given back (excepting
an unavoidable loss) in the form of secondary current, by the
inverse reconstitution of the same products. When the gases
begin to be liberated, it is a sign that the battery is no longer doing
useful work towards producing the secondary current. Thus the
sign of the gases being set free in a secondary cell, previously well
formed, becomes an indication of the maximum charge that the cell
can take, and there is no further advantage gained by prolonging
the action of the primary current, as far as regards the secondary
effects. The cell must be in this condition of previous electro"
chemical preparation, in order that the liberation of gas may
indicate that it is fully charged ; for, with a new secondary couple,
submitted for the first time to the action of a primary current, or
with a couple already formed, but which has remained a long time
without being used, we see the gas set free at the surface of the
plates, almost immediately, without having attained the maximum
degree of charge which they are able to take.(I)

(i) New lead plates exhibit, however, to a certain extent, these signs of the absorbtion
of gas, after the direction of the primary current has been changed two or three times only.
They pass successively through the lightest and darkest shades of peroxide of lead, or attain
metallic tints of a silver grey, according to the gas which acts upon their surface. The gases
are absorbed for the short time corresponding to the developmen: of these thin layers of oxide
and reduced metal,

48

61. Maintenance of secondary cells. — When a secondary

cell is considered sufficiently formed, intervals of rest of several
months, far from being useful, as in the operation of formation,
would tend to increase the resistance of the cells, and to render
their charge longer and more difficult. It is consequently preferable
to charge them from time to time, or to keep them constantly
charged by a weak current, so as to avoid the production upon the
surface of the positive plate, of a layer of protoxide of lea$, of
low conductivity, arising from the slow and spontaneous production
of the peroxide of lead.

62. Secondary elements \\msformed by the current from two
Bunsen cells, may be charged and kept in order by means of a
battery of three Daniell or Callaud elements filled with pure water,
such as those used in telegraphy. The charge obtained, it is true,
is not so strong as with nitric acid elements, but the use of this
primary source is more convenient in a great number of cases.

63. Effects produced by secondary cells with lead

plates. — Temporary calorific and magnetic effects may be
obtained, with the secondary cells that we have just described, far
more intense than the primary battery used to charge them could
produce. We shall show their application in the second part of
this work. The apparatus shown, (fig. 12) (45) has been arranged
specially with a view to demonstrate calorific effects, and we
have already mentioned one or two of these effects (40-41-56).
By coupling up four or five of these cells in parallel, thick wires of
iron or steel may be melted, and molten globules may be obtained
7 to 8 millimetres in diameter. In order to form the globules, the
binding screws of the apparatus must be drawn together gradually
as the wire begins to melt.

49

64. We have made a special study of these globules, on
account of some remarkable analogies which they seem to
show, and which we shall point out in the fourth part. By
examining them with the help of a magnifying glass, darkened,
whilst they are in a state of fusion and throwing out a brilliant

lustre, under the influence of the powerful .el^rtrir nnrrpnf niccinn-

Fig. 13.

Fig. 14.

it MHU.U uuuy oi matter, urougm 10 a very nign
temperature, in which the calorific movement, itself arising from
the previous electrical motion, is not immediately destroyed;
the chemical results of the rise in temperature (such as the

50

oxidation of the carbon, if the wire is of steel) follow also for an
instant the passage of the electric current, and are shown by the
gaseous bubbles, seen upon the surface of the globules.

No doubt these effects could be produced by any other source
of dynamic electricity, in sufficient quantity; but we describe
them here as being obtained with secondary cells, in an easier and
more convenient manner for study, than by any other means.

65. Magnetic effects. — The magnetic effects produced by
discharging secondary cells are also very powerful. Electro
magnets wound with thick wire may be more strongly excited in
this way than by ordinary batteries in which there is much less
surface, and which, consequently, could not afford, in a given time,
so great a quantity of electricity. Powerful permanent magnets
may be easily formed, — so much the more, as, in this case, the
current only needs to be passed for an instant through the coil
surrounding the steel bars intended to be magnetised.

Electro-dynamic experiments, in which the conducting wires
have to carry the greatest possible quantity of electricity, may be
successfully repeated with the aid of secondary cells.

If we want to work an electro-magnetic motor, not for a work of
long continuance, but to carry out some experiments, or for
illustration, these cells may be employed with great advantage.
When the circuit of the electro-magnets is not too low in
resistance, the current of the secondary cell is not so quickly
spent, and we have been able, even with secondary cells of much
smaller surface than that represented in figure 12, in a single
discharge, to drive an electro-magnetic motor for more than an
an hour,

51

These cells may be also used for working induction coils, and
carrying out a great number of experiments, by means of a single
discharge.

66. Formation of ozone in secondary cells, and

voltameters With lead plates. — We have noticed, when
using secondary cells with lead plates, that they often gave off a
strong odour of ozone, especially when they were charged in the
reverse direction by means of a rather strong current.

Nonoxidisable metals, such as gold and platinum, used to be
considered as the only ones which could be used for obtaining
ozone by the electroysis of water. By studying voltameters with
lead electrodes, from this point of view, we found that ozone
could be as easily produced with lead electrodes as with platinum,
and even in a stronger degree.'11

This may easily be proved by taking two voltameters, one of
which is made with platinum wire, the other of lead, of the same
length and diameter, and sending through them both, the current
from ten Bunsen elements. By immersing some strips of paper
iodised and starched, in the open tubes placed above the positive
wire in each voltameter, we see them both grow blue, and it may
be observed that the paper immersed in the oxygen of the lead
wire voltameter grows blue more rapidly, and with more intensity
than the paper immersed in the oxygen of the platinum wire
voltameter.

The pungent smell, and the rapidity of the oxidation of silver,
also show an easily noticeable difference. By causing ozoned

(i) Comptes rendus, t. LXIII, p. 181, 1866,

52

oxygen to be liberated simultaneously from two voltameters of
similar solutions of iodide of potassium, the solution submitted to
the action of the oxygen of the lead wire voltameter turns yellow
more rapidly than that which is acted upon by the oxygen of the
platinum wire voltameter, and we find the quantity of iodine
liberated by the ozone from the platinum wire voltameter is only
about two thirds as much as is given by the ozone from the lead
wire voltameter.

The lead wire voltameter should be slightly formed beforehand,
like the secondary couples (53), so that the layer of peroxide of
lead, produced at the positive pole, thoroughly covers the metal.

67. This appearance of ozone, more abundant with lead
electrodes than with those of platinum, is a rather difficult fact to
explain in the present state of our knowledge of ozone.

Still we think it may be explained by taking into consideration
the influence which the greater or less metallic condition of the
electrode must possess over the generation of ozone. We know
with what readiness metals, or oxidisable materials in general,
absorb ozone; how silver which is not tarnished by the atmosphere,
blackens under its influence.

We may then presume that platinum itself is not absolutely
free from action upon ozone, in proportion as it is generated in a
voltameter with wires of this metal, and that a certain part of the
ozone generated can be destroyed, whilst, in a voltameter with
lead electrodes, the positive, once well coated with insoluble
peroxide, in acidulated water, forms a less metallic electrode than
platinum, and is not so favorable to the destruction of ozone.

53

This appears to be proved by the fact, that in comparing the
quantity of ozone produced by wires and plates of lead, we have
obtained more ozone with wires than with plates, and, by using at
the positive pole of a voltameter, a very short and very fine point
of lead, we have had more ozone than with a thicker wire of a
moderate length. We may then take it, that the less the surface
presented by the electrode, the greater is the quantity of ozone
produced, so that, if it were possible to decompose water without
an electrode, or with as non-metallic or non-oxidisable electrode
as possible, we should have the maximum proportion of ozone in
a voltameter. The formation of ozone by the simple influence of
static electricity, as described by Messrs. Fremy and Ed. Becquerel,
or by the flow through the induction coil, in the tubes of Messrs.
Siemens, Houzeau, de Babo, Boilot, Thenard, Berthelot, etc.,
would seem to support this theory.

68. Shades or tints of oxide produced at the positive
pole during the discharge of secondary cells.— We have

remarked above, when studying the voltameter with lead electrodes
(18), that the production of the peroxide of lead, formed upon
the positive electrode, under the action of the primary current,
was not the only chemical action produced, nor the sole source of
the E.M.F. of the secondary current; but that the other lead
electrode became oxidised at the same time, in consequence of
the decomposition of the water, in the interior of the secondary
cell itself.

This oxidation, hardly visible in a voltameter, is clearly shown in
secondary cells, by a very distinct phenomenon, during the
discharge.

If, for example, we discharge one of these cells, by reddening a
platinum wire, the negative plate at first .-preserves, in the exterior

'HIT 17 BE SIT 7'

5*.

54

visible part, the clear grey tint of metallic lead, during nearly the
whole time that the incandescence lasts; but, upon the wire
ceasing to be red, a dark shade is seen to appear which covers the
exterior surface of the plate and gives it a darker grey tint. The
oxidation of this plate, by the interior current of the secondary
cell, is not sufficiently complete, nor sufficiently prolonged to give
it the shade of peroxide of lead ; but its change in physical aspect
is nevertheless appreciable, and reveals the chemical phase
produced. During the greater part of the discharge, the oxida-
tion takes place inside the coil; towards the end it extends
gradually over the entire plate and, naturally, the outside which
has not the other electrode opposite to it, is the last to be affected.

69. We also notice, especially in the earlier periods of their
formation, the liberation of gas, which is shewn in certain

voltameters (13) upon breaking the primary current. In this case
the phenomenon is still more marked, especially if we close the
secondary circuit with a short thick wire of good conductivity, on
account of the intensity of the current which is generated in the
interior of the cell.

70. Duration of the discharge in secondary cells.—

The length of the discharge of secondary cells is in proportion to
the degree of formation which they have attained. Thus, one of
these couples will maintain a platinum wire, one millimetre in
diameter, at a red heat for from i to 10 minutes, according to its
degree of formation. But, taking the same couple, the duration of
the discharge of course depends also upon the resistance of the
exterior circuit. With a secondary cell which would give, by using
a thick platinum wire, only an incandescence of a few minutes,
we can keep a platinum wire -5 of a millimetre in diameter,
incandescent for an hour.

55

\

The length of the discharge of secondary cells consequently
depends upon the extent of their surface, the thickness of the
deposits on the plates — especially the layer of peroxide of lead,
which penetrates the positive plate — and, lastly, the resistance of
the outer circuit.

71. Constancy in the secondary current during

the discharge.'1' — The resistance opposed to the passage of
the secondary current, in which the quantity of electricity has
more importance than the tension, plays the part of moderator or
regulator, and thus transforms the effect of a current of a
temporary nature, into a relatively constant current of a long
duration. Thus with a resistance of 50 metres of copper wire i
millimetre in diameter, placed in the circuit of a secondary couple
and a tangent galvanometer, the coil of which had a resistance
equal to that of three metres, we obtained a practically constant
intensity of current for about an hour.

72. This constancy on the part of a current, which one would
suppose on first thought must continually decrease from the time
the circuit is closed, is explained by the great quantity of electric-
ity stored, under the form of chemical work, in the secondary cell,
with a relatively low E.M.F., just as a very broad tank,
containing a great quantity of liquid of very little depth, would
give, for a long time, a nearly constant outflow through a small
orifice, and cease quickly when the liquid fell to the level of the
orifice, so, a secondary couple of large surface, whether it reddens
a wire or is discharged through a galvanometer, only shows a
falling off in intensity some moments before completely ceasing
to supply a current.

(i) Les Mondes, t. XXVII, p. 474. 1872.

56

The fact is startling when we discharge the secondary cell
through a fine platinum wire. The incandescence is maintained for
a long time at a uniform degree, and ceases almost abruptly when
the store of chemical work accumulated in the cell is exhausted.

73. We have proved this in a still clearer manner by tracing the
intensity curve of the secondary current, during the length
of the discharge effected through a constant resistance. The
curve presents, for the greater portion of its length, nearly a
straight line, parallel to the base upon which the time is marked,
only falling abruptly towards the end of the discharge.

74. We except here the effect produced during the first
moments of the discharge following the break of the primary
current. Immediately after this break, there is always a maximum
effect, owing to the double origin from which the secondary
E.M.F. arises. This force, as we know from studying the volt-
ameters, (28, 31, 34,) is due to chemical actions exercised by the
primary current upon the electrodes and the liquid surrounding
them at the same time. The products resulting from this latter
action, such as oxygenised water, formed in a very small quantity,
very uncertain, and not very adherent to the electrodes, are
immediately reduced, or dispersed in the remainder of the liquid.
Then, even when the secondary circuit is left open, their action
disappears, as we shall see further on, when treating of the E.M.F.
of secondary couples.

On the other hand, the products resulting from the primary
current on the electrodes, are formed in a certain quantity, remain
fixed to the electrodes, and only alter when the secondary circuit
is closed.

57

s

Hence, two actions help to produce the secondary current in
the cells in question : one only acting during the first moments
following the break of the primary current, the other being
prolonged for an hour.

It is this latter effect — this continuous current from the second-
ary cells — which manifests the constancy noticed above.

75. Preservation of the charge taken by secondary

cells. — Lead plate secondary cells acquire a valuable property by
the work of formation: that of preserving a great portion of their
charge for a considerable time after the action of the primary
current.

Thus, a secondary pair, vte\\ formed and thoroughly charged, will
raise a platinum wire, a millimetre in diameter, to a red heat for
several minutes, two or three weeks after having been charged.
We have even obtained this effect sometimes, with cells except-
ionally well formed^ more than a month after having subjected
them to the action of the primary current.

76. If the peroxide of lead deposited on the positive plate
did not tend to reduce itself spontaneously in the acidulated water,
by local action with the metal beneath it, the preservation of the
charge held by the secondary cell ought to be unlimited. But
this peroxide is reduced, and with so much the more facility as the
layer is thinner, and in a new secondary couple the charge'1* cannot
be preserved at all.

(i) We employ here the word charge, for want of a more exact term, in order to desig-
nate the effect resulting from the accumulation of the chemical work by the primary battery
in the secondary couple. No doubt there may also be a veritable static charge, as in a
condenser, but quite unimportant, in spite of the large surfaces, on account of the conduct-
ivity of the electrolyte, which separates the two metallic plates (35).

58

How is it then preserved, with very little loss, in a formed
secondary cell ?

It may be explained by taking into consideration that, if the
deposit of peroxide of lead has a certain thickness, the surface
in immediate contact with the solution is alone reduced to a state
of protoxide, and then protects the underlying layers from any
action.

77. It seems difficult, at first sight, to imagine a non-conducting
coat of oxide preserving an electrode from 'chemical action when
there is no current passing through the apparatus, and powerless
to protect it when there is an electric circuit formed. But we
know, on the other hand, how powerful is the intervention of
electric energy to modify, or determine, actions which could not
take place without it, and we will quote, en passant^ an instance in
which an analogous effect may be easily set forth.

If we employ, for example, a voltameter with acidulated water
in which the positive pole is formed of a bed of mercury, and the
negative pole by a platinum wire, there is produced at the surface
of the mercury, from the first moments of the passage of the
principal current, a layer of insoluble sulphate of mercury which
soon weakens, in a marked manner, the intensity of the current.
If we reverse the current, this superficial coating no longer offers
any opposition to the electro-chemical action; the sulphate of
mercury is immediately cleared from the surface of the mercury
as if it were swept by a draught of air; it becomes soluble in the
liquid, whilst the surface of the mercury becomes brighter and
soon gives off a steady liberation of hydrogen.

In a lead plate secondary cell the operations take place in the
same manner, when the primary current is sufficiently strong

59 .

and when it has just been reversed. The deposits are sometimes
loosened, and fall in flakes to the bottom of the liquid.

We then conclude that, when the secondary circuit is closed,
the electrical actions called into play cause chemical re-actions,
such as reduction or oxidation, beneath the non-conducting
coatings which act as a protection when an electric circuit is not
made.

78. Residual charge afforded by secondary cells.—

Secondary couples, when discharged, are able to give at the
end of a certain time, without having been again charged, a
residual charge, similar to that obtained from Leyden jars.

If for example we raise a platinum wire to a red heat by the
discharge of one of these cells, and if the circuit be broken
immediately the wire has ceased to be red, we can, in a quarter or
half an hour, again observe an incandescence of several seconds,
by closing the secondary circuit.

If the cell has been very well charged, and if the discharge has
lasted a long time, we can obtain, the next day and several
following days, a residual discharge, the duration of which may be
from two to three minutes.

One could even still obtain a further series of successive
discharges, decreasing in intensity. This phenomenon arises from
the fact of the layer of peroxide of lead upon the positive plate
not being reduced throughout its thickness by the first discharge
of the secondary cell.

In fact, during this first discharge, whilst the peroxidised plate
is being reduced, the other plate becomes oxidised, as we have
described (50), and tends to produce an opposing current in the

60

interior of the pair itself. The E.M.F. of this current, which
may be distinguished by the name tertiary, ends by equalling the
E.M.F. of the pair; the two plates are soon found in an electro-
chemical state, nearly identical, and the discharge of the couple
appears terminated. But if the circuit be again broken, the thin
layer of oxide formed during the discharge, upon the lead plate
previously negative, is reduced little by little in the acidulated
water, as we have also explained above (76). At the end of a
certain time (more or less long according to the duration of the
discharge of the secondary cell), this layer is completely or
partially reduced, and, as on the other hand, the thick coat of
peroxide of lead developed upon the positive plate by the primary
battery has not been entirely reduced throughout its depth
during the discharge, the two plates are found to be again in a
different electro-chemical condition, and consequently, a fresh
discharge may be obtained by closing the circuit. The successive
residual discharges, decreasing in intensity, afterwards obtained,
may be explained in the same way.

79. Increase of the intensity of a secondary cell,

With rest, after being" Charged.— A still more remarkable
phenomenon, sometimes presented by secondary cells, is the
following : —

We take a cell which has been left a long time without being
recharged. As has been already described (61), these cells do not
recharge easily. We send through it for several hours the current
from two Bunsen elements, and when we find that the cell will
heat a fine platinum wire, without being able to redden it, the
primary circuit is broken. At the end of twenty hours rest, it is
noticed that the cell raises this same platinum wire to a red heat,

61 .

Thus, the cell appears to attain, by repose an E.M.F.
higher than that which the primary battery was able to give it.

We think that this fact may be accounted for as follows : —
during the time that the primary current passes through the
secondary cell, the electrolytic gases tend to set themselves free
between the metal and the layers of oxide, — more or less reduced, —
which cover it. These gases not being able to escape easily,
increase by their presence the resistance of the secondary couple,
by preventing contact of the liquid with the surface of the plates,
in proportion as they are oxidised or reduced by the primary
current. If we suspend the action of this current the gases are
slowly liberated, the metallic surfaces changed by the primary
current are brought into better contact with the liquid, and the
secondary cell is able to give, in this quite exceptional case, a
discharge of a greater intensity after the repose than immediately
after the action of the primary current.

80. The Electro-motive force of lead plate second-
ary cells. — The E.M.F. of lead plate secondary cells is
somewhat easier to measure than that of a voltameter, because
the duration of the discharge is longer, on account of the greater
surface of the couples, and the larger quantity of deposits
accumulated. In all cases, there is, as we have seen (74), a
maximum of E.M.F. immediately following the break in the
primary current, which is of rather short duration on account
of the peculiar nature of the chemical actions which produce it.
It is then necessary, for measuring this maximum, to operate in
nearly the same way as with voltameters (6), that is to say, to close
the secondary circuit as quickly as possible after breaking the
primary circuit, and to examine the first effect produced upon the
galvanometer,

62

We have carried out this measure in various ways, either
by means of a tangent galvanometer and changeable resistance,
or with the electro-magnetic scale (9), in operating upon a
single secondary couple or upon a large number connected
in series.

In one experiment among others, made upon 40 secondary
elements, all charged at once, as will be seen further on (ch. Ill),
by three Bunsen cells, coupled up in series, at the moment of
discharge, we have obtained an attraction of the electro-magnetic
scale equal to 9*45 grammes, which corresponds to o'236-gr.
per secondary cell. The E.M.F. of one Bunsen element,
measured by means of the same scale, was found equal to o*i64-gr.

If this E.M.F. be taken as a unit, that of the lead plate
secondary cell will be, by deduction, i -44. In operating upon a
single well formed secondary cell, there is necessarily a more
perfect charge, even with two Bunsen elements as primary source
instead of three, as in the preceding experiment, and most of the
figures we have obtained are comprised between 1-45 and i'5o.

We may, then, consider the electro-motive force of lead plate
secondary cells, observed immediately after breaking the primary
current, as about equal to one and a half times that of a Bunsen
element, and about two and a half times that of a Daniell element.

It is, besides, the result which we obtained with a simple
voltameter (20).

81. If we measure this E.M.F. two or three minutes after
breaking the primary current, then, even when the secondary circuit
has remained open, it is found to be noticeably diminished, and
reduced to 1-17, by reason of the causes which produce a
temporary polarisation, as previously mentioned (74), disappearing.

63 >

But the E.M.F. is maintained very constant at this point
during nearly the whole of the discharge.

82. This difference between the initial E.M.F. of a lead plate
secondary cell, and its subsequent E.M.F., is, besides, easily
appreciated when a secondary couple is discharged through a
platinum wire immediately after the action of the primary battery.
For the first moment the incandescence is very high, almost
fusing the wire. If we allow, on the other hand, an interval of a
few minutes between the action of the primary current and the
discharge, the incandescence is not so high but very uniform
until the end of the discharge of the secondary cell (73).

83. Resistance of lead plate secondary cells. — We

first found out the resistance of the secondary cells'1' by a plan
similar to that sometimes used for measuring the resistance of
voltameters, namely, opposing two secondary cells to each other, by
means of a special commutator at the instant the primary current
is broken, so that their E.M.F. may be neutralised and the double
resistance alone put into play.

But since we have attained, by the means of the formation, a
discharge of some duration and certain constancy, we have
been able to measure the resistance of these cells, previously well
charged, like ordinary constant current cells.

The method of employing a tangent galvanometer and variable
metallic resistances, not having afforded us very consistent figures,
we have latterly given preference to a plan, based upon the use of
derived currents, due to Sir William Thomson, simplified by
M. Mouton,(2) and suggested, with some reason, as one of the most

(1) Annales de Chimie et de Physique, 46 serie, t. XV, p. 19, 1868.

(2) Journal de Physique, par ch. d'Almeida, t. V, p. 144, 1876.

64

convenient and rapid that can be employed, for measuring the
resistance of voltaic cells.

We have found that the resistance of secondary cells of the
various dimensions which we have used, varied from 2 to 5
metres of a copper wire i millimetre in diameter.

We have noticed that the extent of surface, or the size of the
secondary couple, which has a great effect upon the duration of
the discharge, has far less influence over the resistance of the cells,
than the distance the plates are apart, their more or less perfect
degree of formation, and their condition. Thus, cells of very
small surface (2 square decimetres) in which the platesw ere only
separated by a distance of 2 millimetres, had but a resistance of
about 3 metres of copper wire i millimetre in diameter, whilst cells
having half a square metre of surface, the plates of which were
5 or 6 millimetres apart, and which had remained a long time
without use, possessed a resistance of 4 to 5 metres of the same
wire.

However, in any case, these experiments prove that the
resistance of the secondary cells is very low, and the intensity of
effects obtained from them may be thus explained.

84. Necessary E.M.F. of the primary current.— We

may draw from the preceding tests relative to the E.M.F. (80),
that, to charge lead plate secondary cells, it is only necessary to
employ a primary current of a higher E.M.F. than one and a half
times that of a Bunsen element. And two of these elements are
perfectly suitable for the purpose, as we have already seen (41).
Three elements would, no doubt, charge the secondary cell more
quickly; but this excess of E.M.F. is not necessary and may even

65 •

prove inconvenient, if the cells have been formed with a weaker
current, in loosening the coatings of oxide and reduced metal
upon the plates by a too rapid liberation of gas.

If it be desired to employ Daniell elements as the source of
the primary current, three of them are sufficient to exceed the
opposing E.M.F. of the secondary cell; but, as the excess is not
so great as in the case of two Bunsen elements, the secondary cell
is not so completely charged. Besides, they afford, in a given
time, a less quantity of electricity. Also, the charging occupies
very much longer time, and as there may be causes for loss, of
which we will speak further on (92), these elements do not charge
the secondary cells so completely. They do very well, all the
same, for keeping them charged (62).

85. Limit of the charge that secondary cells may

take. — It would seem that, by increasing indefinitely the extent
of the surface in a cell, or by coupling a number of cells together,
one might obtain, with a weak primary battery, a secondary
current of an indefinite intensity. But there is a limit, beyond
which we cannot go, in prolonging the duration of the charge.
Just as we cannot thoroughly charge batteries of condensers (in
static electricity), of large surface, by means of very small
electrical machines, on account of the losses through the
atmosphere when the charge takes too long, so, in this case, there
exists a cause of loss whilst the charging goes on, in the tendency
of the peroxide of lead to be spontaneously reduced in the body
of the acidulated water, whilst it is being formed (76).

This reduction is so much the easier when the layer is more
slowly deposited, and consequently thinner, so that a time comes
when the action of the primary current, in order to renew or

66

maintain this layer upon the surface of the plate, is balanced by
the tendency of the peroxide of lead to be reduced in the liquid.
The limit of the charge that the secondary cell can take from a
certain primary source is thus fixed.

86. Secondary cell charged by a thermo-pile.— Any

apparatus giving a continuous current of electricity in the same
direction may be used to charge secondary cells, provided it has
sufficient E.M.F. as we have just seen (84). Thus, for example,
a secondary cell may be charged by one of Ed. Becquerel's or
Clamond's thermo-piles, and if there be produced, in discharging
it, the incandescence of a platinum wire, a portion of the actual
heat used for the charge, which is stored in the form of electro-
chemical work in the secondary cell, is expended in this manner.

87. Secondary cell charged and discharged by
means of the Gramme machine.— A secondary cell may

be equally charged by means of mechanical work, and the charge
may be reconverted into the same form, any length of time after
having been carried out.

This is the result of an experiment, which we have made
together with M. Alfred Niaudet,(l) with the help of the Gramme
machine, which gives, as we know, a continuous current in the
same direction ; and being reversible, like magneto-electric
machines in general, can serve as an electro-magnetic motor.

Figure 15 represents this experiment. A Gramme machine with
Jamin magnet is coupled up to a secondary cell, which may be, for
the purpose of illustration, of much smaller dimensions than that
described above (45).

(i) An electro-dynamic experiment by Messrs. G. Plante and A, Niaudet. Comptes-
rendus, t. LXXVI, p. 1259, 1873.

67

If, after having charged the cell, the machine be stopped,
without breaking the connection between the two parts of the
apparatus, we see it immediately start in motion under the influence
of the discharge, thus proving a fact which seems at first sight
contradictory : namely, that the machine turns, not in the opposite
direction, but in the same direction it was running when charging
the secondary cell.

This fact is explained in the following manner : — if we first
consider the direction of the current furnished by the machine,
then that of the current given back by the secondary cell (which is
in the opposite direction to the preceding one), and if we take into
account the actions resulting from it, we see that according to the
laws of induction and electro-dynamics, the motion of rotation
ought certainly to be in the direction shown by the experiment.*
On the other hand, it is necessary to mention, that in charging the
secondary cell, the Gramme machine, driven at a high speed,

* We presume it is a dynamo-electric machine and not one with a permanent magnet
which the Author refers to in this case. (Note by Translator.)

68

attains a higher E.M.F. than the opposing force developed in the
cell. When the machine is stopped, and the cell, by discharging,
causes it to revolve in the same direction, the speed given to it is
not high enough to cause it to generate a higher E.M.F. than that
possessed by the cell itself.

The rotation, then, takes place by reason of a difference between
the two opposing electro-motive forces : that of the charged
secondary cell, which predominates, and the weaker one which the
machine tends to develope by its movement, under the influence
of the discharge of the secondary cell.

We may say that, in this experiment, the Gramme machine,
which produces the same effect as a battery immediately it is set
in motion, polarises under the influence of the discharge of the
secondary cell, since it presents an opposing E.M.F., so that we
thus find a mechanical reproduction, so to speak, of voltaic
polarisation.

88. Various analogies presented by secondary

cells. — On taking into consideration the effects produced by
secondary cells, of which we have just exhibited the principal
properties, we find that these apparatus can play the same part
in dynamic electricity, as the Leyden jar and condensers in static
electricity. The analogy continues, even, as we have seen, to the
residual discharges which they give.

Nevertheless, we have remarked (48) that the cause of the
production of the current in the secondary cells was purely
chemical; that if there were condensation of electricity, properly
called, this effect was not worthy of notice, and that these
apparatus did not directly store the electricity itself but the

69

chemical work of the battery. We will not then insist any more
upon this analogy, but will point out a few of another order,

offering no less degree of interest.

^t>-*v^rd

89. These secondary cells may be also used in connection

with all apparatus employed in mechanics, for the accumulation
of work resulting from the action of force, such as hydraulic
accumulators, compressed air reservoirs, springs (so called
secondary motors), pulley blocks, winches, &c., going back to the
simplest appliance, the lever itself.

A secondary couple is, in fact, a kind of lever for dynamic
electricity ; because one can thereby obtain with a weak electrical
power, an increase of this power, in such proportion as one wishes,
on the condition of loss in speed, or a necessary sacrifice of time,
in order to accumulate the effect.

The same principles as those which apply to the lever ought to
be taken into consideration ; otherwise, the secondary cells might
cause both "illusions and labours, the uselessness of which could
be demonstrated in the same way as the impossibility of perpetual
motion.

90. One of the principal advantages presented by these
secondary cells is to afford storage for spare electrical work, or as

,hott^t<O

one sometimes expresses it, at the present time, a powerful energy
that may be used at will, in a longer or shorter time.

We have considered, up to the present, that this power is
expended in a shorter time than was required for its accumulation, so
as to obtain an effect of greater intensity than that of the primary
force. But it may also be interesting, in certain cases, to charge a
secondary cell with a greater force, in a very short time, and to
expend the accumulated work during a more extended period.

70

The following experiment is, from this point of view, very
illustrative.

We take as a primary electric source a battery of two Bunsen
elements, strong enough to raise a platinum wire, half a millimetre
in diameter, to a red heat, and we charge by this means, for one
minute only, a well formed secondary cell. We then discharge
the secondary cell through a considerably finer wire than that
reddened by the primary current, a wire *i millimetre, for
example, and we notice that this wire continues red for about five
minutes.

The expenditure of the stored work has, in this case, lasted
much longer than was required for its accumulation. An effect is
thus realised analogous to that produced by sharply drawing the
cord wound round a child's top, and then letting the toy expend, by
a prolonged rotary motion, the motive force which was com-
municated to it in one brief instant.

91. Return obtained from Secondary cells.— Taking
this view of secondary cells, by comparing them with apparatus
for accumulating mechanical work, leads us to measure their
efficiency; or the proportion which the electrical work restored by
their discharge bears to the electrical work expended in charging
them.

The work most directly effected by the voltaic current being
chemical, we have compared the total of the chemical actions
produced in the circuit during the charge, with that of the same
kind of actions produced in the discharge. As a means of
comparison we chose the reduction of sulphate of copper, it
being the easiest electro-chemical action to measure.

71 .

A sulphate of copper cell, furnished with a platinum plate,
previously weighed, was joined to the primary battery of two
Bunsen elements ; on the other side, a well formed secondary
cell was placed in connection with the primary battery for a certain
time, and the passage of the primary current was stopped immed-
iately the liberation of gas began to appear in the secondary cell,
the latter then being, as we have seen above (60), charged nearly
to saturation. (The platinum plate of the test cell, coated with
copper, was weighed after the experiment.)

This done, we discharged the secondary cell by closing its
circuit through a voltameter of sulphate of copper, provided with
another platinum plate, already weighed, and we did not stop the
experiment until the action of the secondary current was
completely exhausted. We considered this result reached when
the deviation of a galvanometer, also in the circuit, was reduced
to zero.

By comparing, according to the deposits of copper obtained,
the total chemical work returned by the secondary cell during its
discharge, with the total chemical work expended in charging it,
we found that the proportion of return was from 88 to 89 per cent.

92. The loss of work corresponding to the n or 12 per cent
missing in the return may be accounted for as follows :

First, the spontaneous reduction in the acidulated water of a
small portion of peroxide of lead, in proportion as it is deposited
upon the positive plate; a cause so much the more important as
the surface of the secondary cell is greater, and the deposited
layer consequently thinner, and as the charge lasts longer. With
a cell of extremely large surface in proportion to that of the

72

primary battery intended to charge it, this cause of loss would
increase to an indefinite extent, and it would be hardly possible to
charge the cell.

Secondly, the incomplete formation of the secondary cell :
a portion of the gases being then driven off without producing any
useful chemical effect in the ulterior development of the
secondary current.

Thirdly, the polarisation or the development of opposing E.M.F.
in the interior of the secondary cell itself, whilst it is at work.
The consequence is, that when discharging the cell, we lose the
portion of work wasted from this cause. In order to estimate this
loss, it would be necessary to ascertain the work which could be
produced by the residual charge (78), and add it to the return.

However that may be, in spite of these sources of loss, we see,
according to the return obtained — nearly equal to 90 per cent. —
that a lead plate secondary cell, well formed, affords a very perfect
accumulator of the work of the voltaic battery.

CHAPTER III.

Transformation of the energy of the
Voltaic Battery by means of lead
plate Secondary Batteries.

Secondary Batteries of high tension.— Various

arrangements. — Their effects. — Instructions

regarding the use of Secondary

Batteries. — A nalog ies.

93. The results we have just shown allow, as we have seen,
of the accumulation of a quantity of electricity arising from a
given voltaic source, without obtaining a higher tension or E.M.F.
than that of the source. But many electrical demonstrations
require not only a quantity of electricity, but a considerable
tension. It was, then, interesting, to try to obtain, in an easy
manner, and without too much loss in the transformation, some
effects of a higher E.M.F. than that of a given electric source.

74

Grove's gas battery offered the first means of approaching the
solution of the problem. In fact, by charging successively a
certain number of these gas cells, by means of the same primary
battery, so as to fill them with the gases arising from electrolysis,
a battery is formed of a higher E.M.F. than that of the primary
one. But the gas battery being able to give only a very small
quantity of electricity, and each of the cells having, besides, but a
very low E.M.F., this solution of the question presented more
interest from a theoretical than from a practical point of view ;
because it was difficult to take advantage of, even for scientific
research.

The apparatus known by the name of the De la Rive electro-
chemical condenser,*1' permits of the production of the electrolysis
of water in a voltameter with platinum electrodes, by employing
the extra current developed in an induction coil with a single
voltaic cell; a result which could not be attained directly from the
cell itself, and proving in consequence, the development of an
E.M.F. higher than that of this cell.

The works of Poggendorf(2) show that we can obtain, by means
of several voltameters with platinum electrodes, polarised by a
given current, an increase still more marked, and even indefinite,
in the E.M.F. of the current, by collecting, with the help of a
pivoted mercurial commutator, the polarisation current emanating
successively from all the voltameters, connected in parallel and in
series.

(1) Archives de 1'Electricite, t. Ill, p. 159, 1843; et Traite d'Electricite par
A. De la Rive, t. I, p. 391.

(2) Annales de Poggendorf, t. LXf p. 568, 1843 ; et t. LXI, p. 586, 1844.

75

M. J. Mtiller(t) employed, for the same object, a spring commu-
tator, without mercury, which permitted of a continuous rotary
motion.

M. Thomson's polarisation battery'2' also offered a solution of
the problem, by allowing a series of platinum voltameters to be
successively charged with great rapidity, and discharged with the
same rapidity, so as to finally obtain a continuous current of a
higher E.M.F. than that of the primary battery used to charge
them.

94. We have in our turn applied'31 lead plate secondary cells,
as described in the preceding chapter, to the production of a
current of higher E.M.F. than that of the primary battery, by
using their opposing E.M.F., already somewhat raised in itself,
and the persistency of its action after the passage of the primary
current.

Lead offering, besides, the advantage of being easily arranged
with large surface, we have been able to produce far higher effects,
both in E.M.F. and quantity simultaneously, than those of the
primary current, thus transforming and accumulating the work of
the voltaic battery at one and the same time.

95. Secondary battery for tension effects, formed of

parallel lead plates.— Fig. 16 shows the first arrangement we
used. Secondary pairs to the number of 4o'4) were formed, each

(1) Traite de Galvanisme et d' Electro- magnetisme, by G. Wiedemann, and edition,
t. I, p. 657-

(2) Annales de Poggendorf, t. CXXIV, p. 163, 1865.

(3) Annales de Chimie et de physique, 4th serie, t. XV. p. itt 1868.

(4) Only 20 cells are represented, in order to give distinctness to the figure.

76

composed of two lead plates, 0,200. x 0,200. in size, held in very
narrow gutta percha cells and immersed in acidulated water.
Each of the lead plates terminated with little copper plates
which had springs at the two extremities and could be pressed,
either by the metal bars MM', NN', or by an insulated bar BB'
furnished underneath with metal pieces. These bars were
arranged so as to form a frame to which an oscillating motion
could be given.

Fig. 16.

In the position in which the frame is represented by Fig. 16, all
the springs connected with the positive plates are pressed by
the rod MM', and all those which communicate with the negative

77 '

plates are pressed by the rod NN'. The secondary couples are
thus connected in parallel, or for quantity, and are charged through
the medium of the wires HH', from a battery of three Bunsen
elements placed near the apparatus.

When the insulated bar BB' is lowered, the metal portions of its
lower surface thus press and unite, two and two, the springs com-
municating with the adjacent poles of opposite names of all the
secondary couples. The battery is then in series.

By connecting the two terminal springs by means of the wires
GG', to the metal rods provided with nippers, a platinum wire
two metres long and '4 of a millimetre in diameter, can be main-
tained at a red heat for from one to two minutes. (I)

96. Secondary battery for currents of high tension,
formed of couples composed of coiled lead plates.— As

the cells of the battery we have just described presented, in time,
some of those faults above pointed out (42), we replaced them by
pairs of plates rolled into a spiral form, constructed as seen in
Fig. n, and have arranged the apparatus as represented in Fig. ly.'2'

Twenty secondary couples, held in cylindrical glass jars filled
with acidulated water, are placed in two rows, and are connected
to the springs of a commutator similar to the preceding one,
intended to alternately unite the cells in parallel during the
charge, and in series for the discharge.

Two copper cylinders, CC, C'C' are fixed to a bar ot insulating
material (wood or vulcanite) furnished with small metallic plates,

(1) When the secondary couples were new, a platinum wire of '2 of a millimetre in
diameter only could be rendered incandescent, and only for some seconds ; but as the couples
are formed by use, it has been possible to incandesce, for a longer time, wires of double
that diameter.

(2) Les Mondes, vol : 27, p. 427, 1872.

78
so that they may be turned simultaneously in either direction by

Fig. 17.

means of a button B, and rub, alternately with the bar, against the
springs r r r.(t)

The combination of all the secondary cells in parallel during

Fig. 18.
the charge, is represented by the diagram Fig. 18, in which, to

(i) This commutator has been cleverly made by M. J. Morin,

79

simplify it, the cells are shown by two plates. The plates of the
odd row Px P3 Ps and P39 are connected all together, and the
plates of the even row P2 P4 P6 P40 on the other hand, are also
connected together, when the springs rub against the metallic
cylinders, here represented by two simple lines, to which are
attached the wires from the primary battery.

The plates of the even row thus become all positive, during the
charge, and those of the uneven row negative. The combination
of the secondary couples in series for the discharge is represented
by Fig. 19.

Fig. 19.

When the springs rub against the metallic parts of the insulated
bar, all the couples are connected by their poles of opposite names
and the discharge can be taken from the two end plates, the direction
being opposite to that of the primary current, as indicated by the
arrows in the two figures.

Fig. 17 (see p. 78) represents the battery producing a voltaic
arc, the commutator being turned into the position which it
occupies during the discharge.

To recharge the battery, the commutator must be turned a
quarter of a revolution. All the cells, then connected so as to
form but one of very large surface, are submitted to the action of

80

the primary current, supplied by two Bunsen elements, the poles
of which are connected with the terminals I and I'.

97. The coiled plates of the cells in the battery are 12 centi-
metres broad and 18 long. The space between the plates is from
3 to 4 millimetres and their useful surface is about 8 square deci-
metres. The resistance of each of these couples when charged is
equal to 8 metres, 77 centimetres of copper wire, i millimetre
in diameter, or about equal to that of a Bunsen cell of the same
surface.

But, as we have said previously (83), this resistance may vary
considerably according to the degree of formation of the plates,
and the distance which separates them. Some couples of even
smaller dimensions may present less resistance.

98. Figure. 20 represents a secondary battery of smaller size
than the last, and of a more simple construction, which we finally
adopted for investigating the effects of electric currents ot
high tension.

The commutator is simply a wooden bar, on the edges of which
are placed bands of copper, and metal pins pierce the wood at
right angles. There is some similarity between this arrangement
and the commutator of Cooke & Wheatstone's needle telegraph.

In the position of the bar shown by Fig. 21, the longitudinal
copper bands gg>, seen in section, touch all the springs such as rr',
and unite all the couples in parallel ; the metallic pins, one of
which is represented by hh', and which traverse the bar, are insulated
from the circuit. In the other position of the bar (fig. 22), the
pins as shown at hh', touch the same springs rr\ and unite all the
couples in series. The surface of the small couples of this battery

81 .

has been reduced to about two square decimetres, so as to take
less time in charging them. The lead plates of which they are
composed are in fact only 10 by 6 centimetres.

Fig. 20.

But as they are very near together, the resistance to the conduct-
ivity of these couples is very small. The commutator is
represented, fig. 20, in the position it must occupy during the
charge of the battery.

Fig. 21. Fig. 22.

The binding screws QQ' communicate with the longitudinal
copper bands of the bar and serve to redden or fuse short thick
wires when the secondary couples connected in parallel have been
submitted for a short time to the action of the primary current.

82

Similar binding screws may be adapted to the commutator
cylinder of the battery already described: (96), fig. 17. The
binding screws TT, fig. 20, join the outer poles of the battery, and
allow the long fine wires to become incandescent or melted, when
the commutator is turned so as to unite all the couples in series.

99. Effects produced by secondary batteries composed

Of lead plates. — We have just quoted, in describing each battery,
some of the effects which can be produced ; we will further add,
that the duration of these effects depends on the more or less
complete formation of the secondary couples which compose them
(53). The potential of the current which they can supply of course
depends on the number of couples. The E.M.F. of each cell being
equal, as has been seen (80), to about one and a half Bunsen ele-
ments, and as, on the other hand, the resistance, with equal surfaces,
is practically the same as that of the Bunsen element (97), there
may be obtained, with a battery of forty couples, or with two
batteries of twenty couples united, the same effect during the
first instant of the discharge, as with a Bunsen battery of about
sixty elements.

Experiments which take but a short time may often be repeated
with a single charge. We may mention, among others, the melting
of a steel wire, which can be effected, in a length of i metre 20
centimetres, with batteries of 40 couples, or i m. 6 cent, with 20
couples.

It is noticed that this fusion is accompanied by the formation of
a chaplet of little molten metallic globules, visible when looking
at the wire through darkened glass, or when examining fragments
of the wire, broken after fusion. Fig. 23 represents this effect
produced with the little secondary battery of twenty couples, last

83 .

described, and we shall have occasion to refer to this experiment
(IVth part) in order to account for the appearance presented some-
times by certain natural phenomena.

Fig. 23.

Without speaking here of effects which we describe further on
and which are obtained by a large number of batteries united, we
will just mention, among experiments which may be repeated with
secondary batteries, that of light produced by the vaporization of
mercury.

By placing a metal cup containing a few drops of mercury, in
connection with one of the poles of a battery of twenty or forty
couples, under a cover through which passes a rod terminated
by a platinum wire, there is obtained, by bringing this wire
in contact with the mercury, a beautiful light, which may be
prolonged for three minutes.

Electro-chemic reactions which require a current of considerable
tension, such as the decomposition of potash or sodium, may be
demonstrated by means of these batteries.

100. Large secondary battery of extended surface for

effects Of quantity Or tension.— In taking secondary couples

84

of large surface, such as that represented in fig. 12 (45), each being
half a square metre in area, instead of cells of 2 to 8 square deci-
metres, and by arranging them as above, effects of quantity
and tension are obtained simultaneously which last a considerable
time, and thus is realised the simultaneous accumulation and
transformation of the work of the voltaic cell.

With six large elements, well formed and arranged in the large
surface battery represented in fig. 8 (38), but with a commutator
placed above, so as to be able to unite them in series, after the
charge in parallel, a voltaic arc is obtained which lasts from seven
to eight minutes, endowed with greater brilliancy than that from a
battery of equal E.M.F. formed of Bunsen elements of ordinary
dimensions.

101. Secondary battery of lead plates for prolonged
effects Of tension. — We have also arranged'1' a secondary
battery of lead plates, composed of 40 elements of very small
surface (a few square centimetres each), for obtaining continued
effects of tension, for which we will point out a purpose
farther on (123).

In this case, there is only a simple transformation of the work
of the primary battery, without accumulation. Tension is pro-
duced at the expense of quantity of electricity, and it is then
better that the two Bunsen elements destined to charge the battery
should be of rather large size.

We will limit ourselves to mentioning this apparatus, described
in the memoirs quoted, intending to modify the arrangement in
order to simplify the application.

(i) Annales de Chimie et de Physique, 4th series, vol. XY, p. 27, 1868,

85

102. Instructions respecting the use of secondary

batteries. — To obtain the maximum effect that the secondary
batteries previously described can give, it is well to be sure that
each couple is in good condition, that is to say, sufficiently 'formed,3
that there is no interior contact between the plates, and that the
terminal tongues which serve as poles are not broken.

The effect produced by each couple of the battery previously
charged may be tried separately and quickly by first turning the
commutator a little, so that the springs joining the poles of all the
couples are not in contact with any metallic part, and then by touch-
ing the poles with the two small copper plates of a little rheoscope
(fig. 24) formed of a platinum wire stretched between two
separate nippers. If this wire is from *2 to -3 of a
millimetre in diameter, and 4 to 5 cent, in length, each
couple of the battery ought to redden it brightly for
about a minute.

If a couple should not redden it at all, this couple
Fig 24 ought to be examined to find out if there is an interior
contact or one of its poles broken.

103. The case of interior contact or short circuit rarely happens
if the couples are well made. It is besides easily recognised, for
contact in a single couple would prevent all the battery from
charging, by presenting much less resistance to the primary current
than the other couples where the plates are separated by
acidulated water. If, then, other couples of the battery are found
charged, it is certain that the couple suspected has no interior
contact.

It may be further proved by touching, for a few moments, the
springs in connection with this couple with the two extremities

86

of the wires from the primary battery. If there is an escape of
gas, the couple has evidently no contact. If there is no escape of
gas, there may be contact in the couple, or it may be broken.

In this case, the spark on breaking the circuit of the primary
battery should be examined, the secondary couple being in the
circuit. If this spark equals in vividness that which the battery
alone would produce, there must be an interior contact. If there
is no spark, it indicates, coinciding with the absence of escape of
gas, that there is a pole broken. (I)

104. The fracture of a pole is the accident which is most
likely to happen to secondary cells of lead plates.

The metallic porosity which plays so large a part in the operation
of the formation of couples, by allowing the electro-chemic action
to work to a certain depth, has, on the other hand, the disadvantage
of allowing acidulated water to penetrate to the interior of the
small terminal plates of the cells, and to gradually advance, until
it comes in contact with the wires or small copper plates which
bind the couple to the springs of the battery. Copper is attacked
especially when in presence of another metal, and the connections
are broken or impaired.

The creeping of the solution is further facilitated by the pheno-
menon of "electric carrying"'2' during the discharge, of the same
kind as that observed by MM. Reuss, Porret, Becquerel and G.
Wiedemann, which works from the positive to the negative pole ;
and it may here be remarked that the accident we speak of happens
nearly always to the positive terminals of the secondary couples.

(1) However, if the battery is weak and the couples have remained a long time without
working, so as to offer, at first, rather a high resistance, it must not be hastily concluded
that the couple has a pole broken ; one must wait some time after the current has passed,
and try if a secondary current is produced by means of the rheoscope of platinum wire.

(2) Known as "Porret's" phenomenon." (Translator.)

87

These small positive plates are also liable to break at the level of
the acidulated water, doubtless on account of local action
between the immersed portion of the peroxydised lead plate
and the exterior portion less peroxydised. In short, after the
discharge of the secondary couples in a circuit of very low
resistance, these terminal tongues may become powerfully heated,
and thus acquire extreme fragility. In experiments we have
made with numerous batteries, of which we shall speak further
on (III. part), this heating of the terminal tongues of those
couples which are not so well charged as others, and submitted
to the action of all the other couples, may reach incandes-
cence and fusing of the plates with a kind of explosion
arising from the rapid vaporization of part of the liquid in
the cell.

105. This accident may be prevented by using tolerably thick
lead (i millimetre in thickness), and by varnishing the tongues of
lead, whilst hot, considerably below the point where they emerge
from the liquid (45).

When this accident happens, it may be remedied by opening
and washing the secondary cell, and by shaping another tongue
out of a portion of the lead plate which was immersed
in the liquid. The tongue thus cut, must be bent round, varnish-
ed, scraped at the end, and fastened by a little binding screw or
nut, either varnished or enveloped in a bed of cement, on to
the wire meant to connect the secondary couple to the terminals
of the battery.

106. If it be desired to throw a couple, to which an accident has
happened, out of circuit, without detaching it from the apparatus,
it is only necessary to unite the screws of the springs which are

88

next each other by a bridge formed of a very small copper plate
with the ends rasped like forks. In this way the circuit presents
no interruption when the apparatus is discharged in series.

107. When there are found, among the couples of a battery,
some which become less charged than others, in consequence
of inevitable variations in the resistance of these different
couples, they are thus known :• — During the discharge, instead of
contributing to the development of the secondary current, they
act, relatively, as simple voltameters, and only give an abundance
of gas under the influence of the secondary current produced by
the other cells. These couples become equal with time, but
if they are wanted to be charged separately and to be more formed,
so as to put them on a level with the others, it suffices, without
detaching them from the apparatus, to place a band of paper
between the springs of the other couples, and the corresponding
metallic portions of the commutator, so as to allow only those
couples to be charged. These couples, if too numerous, would
hinder demonstration of the effects of the battery, for they
would absorb a part of the force of the secondary current during
the discharge; it is necessary then for the battery to be equalized
and this comes with time if the battery be frequently worked.

108. Returns.— Comparisons. —The secondary batteries
above described, being more especially arranged to produce
effects of tension, afford a return inferior to that of batteries
intended for effects of quantity, and less capable of accurate
measurement because of the differences of resistance of the
various couples of which they are composed. These are not how-
ever, the less efficacious as instruments of transformation, which,
after the action of a weak current, permit of obtaining for a
considerable time, the most intense effects of the voltaic battery.

89 *

They may be compared with several machines used in
mechanics for transforming and accumulating force, particularly
the machine known by the name of the "pile driver." Indeed,
in this latter machine, a heavy weight, raised gradually to a great
height by a series of successive efforts, is let drop, and performs
by its fall, in one instantaneous blow, the principal part of the
work done during a considerable time.

In the secondary batteries of which we are speaking, the amount
of chemical action produced by a weak current of electricity,
distributed over a great number of couples, developes an amount
of electro motive force, which, united upon the closing of the
secondary circuit, gives back, in the form of a very intense current
of short duration, the amount of work accumulated during the
time that the charge from the battery lasted. The effects of
quantity correspond with the fall of a very heavy mass, only raised
to a slight height ; effects of tension are analogous to the fall of a
mass less heavy raised to a great height.

These comparisons also show the connection which exists between
divers manifestations of force, or of motion in general, and the
variety of effects which may be obtained, by analogy, with electric
force.'1'

(i) On the employment of secondary currents for accumulating or transforming effects
of the voltaic pile. Comptes rendus, vol. LXXIV, p. 592, 1872.

PART.

APPLICATIONS.

Uses in Galvano cans tics and Therapeutics in
General; — Exploding Mines; — Domestic Uses; —
Electric Breaks; — Signalling by means
of Lights; etc.

109. GalvanocaustiC applications.— The calorific effects
of the secondary cells, above described, may be utilised in galvano-
caustics, for operations which do not require an action of long
duration, and such cases often occur in therapeutics. We pointed
this out in i868(I) and put it into practice in 1872, after finding
that it was possible to increase the charge of secondary cells by
the process of "formation," and also to endow them with the
property of preserving the charge for a considerable time without
great loss.

(i) Researches in secondary currents and their application.— Annales de Chimie et de
Physique, 46 serie, t. XV, p. 21.

91

Fig. 25 represents the arrangement of a secondary cell for this
purpose. The cell is enclosed in a box, the top of which is fitted
with metal plates connected to the two poles where the conductors
of the cauterising apparatus terminate. The essential portion of
the latter is a platinum wire doubled into a point or twisted,
according to the object required to be effected.

Secondary cells arranged in
this way, once charged, may be
easily carried about, and will afford
without any handling in the
presence of the patient, the heat
necessary for the operation.

If the operations take only a
short time, the store of electrical
energy in the secondary cell suf-
fices for carrying out several of
them without recharging.

Cauterizations of the lacrymal
gland have been effected by
Dr. Ominus, in 1873, upon seven
or eight subjects successively,
without it having been necessary
to recharge the apparatus.

When one of these cells, of the dimensions above described
(45), has been well "formed," it will raise to incandescence a
platinum wire i m.m. in diameter by 7 or 8 centimetres long, for
eight or ten minutes, and a wire of the same length and half a
millimetre in diameter for more than twenty minutes.

110. For small operations, such as occur in dental surgery, a
smaller cell may be used, still more easily carried, such as that
shown in fig. 26.

These cells are of the same dimen-
sions as those composing the battery
shown in fig. 17 (96 & 97). They
may be completely sealed by means
of a small plug of india rubber fitted
in the glass tube which passes
through the stopper of the cell, shut
up in a case and easily carried in
the pocket.

Fig. 26.

These cells, when well formed,
will redden a platinum wire half a

millimetre thick for two or three minutes, and a wire '2 m.m.
thick for five or six minutes.

Dr. Moret used these little cells, with success, in the treatment
of neuralgia by means of cauterisation, of the kind known as trans-
current cauterization, and for instantaneously stopping arterial
hemorrhage.*1'

111. This latter type of secondary cell (fig. 26), with only a
comparitively small surface, may be kept sufficiently charged by
means of three Daniell elements. For cells of larger size (fig. 25)
this source would be too weak and we recommend two Bunsen cells
with the zinc in pure water, instead of acidulated, and the carbon
surrounded by nitric acid in a porous jar. Such elements are, of
course, weaker than those made up in the ordinary manner ; the

(i) Revue de Therapeutique, 446 annee, p. 172. Avril 1877.

Among the physicians who were the earliest to interest themselves in this application and
who made use of these secondary cells in their operations, we are happy to mention the
names of Drs. de Bonnefoux, Lailler, Lubinoff, Moretin, Constantin Paul, de Tavel, etc

93

zinc is only attacked slowly and feebly by the acid which finds its
way through the porous jars ; but the trouble of amalgamation is
avoided, and these elements will keep in condition for charging
secondary cells for about a week without it being necessary to
dismount and renew them every day. More time only (about four
or five hours), is needed, when charging with these cells, than
when Bunsen elements of amalgamated zinc in acidulated water
are used.

It is essential to always charge secondary cells in the same
direction, as we have already stated, that is to say, to connect the
positive pole of the primary battery with the peroxidised lead
plate, and the negative pole with the lead plate preserved in a
pure metallic state or coated with reduced pulverulent lead.

It is also of importance to keep the secondary cells open during
the charging and only to entirely close them when carried about;

for the gases given off during
electrolysis would gradually
cause a considerable pressure
upon the liquid if the cells
were closed, and would tend to
make it creep along the lead
terminals or exude by the
smallest cracks in the stopper
and thus affect the points of
contact with the copper strips
Fig. 27. to which are fitted the

cauterizer conductors.

Omission of this precaution is a frequent cause of deterioration
in the connections of secondary cells.

94

112. Use in lighting dark cavities in the human
body and obscure hollows in general.— If the discharge

from a secondary cell be made to pass through a platinum wire
doubled at the middle into a point, as shewn in fig. 27, the wire,
being less quickly cooled, in this form, by the surrounding air,
attains a degree close upon fusion of the metal, and it emits, by its
incandescence, a very bright light which may be made use of.

In various conferences upon the effects created by these
secondary cells, held in 1872 and during the following years, we
had an opportunity of lighting the assembly for from half an hour
to an hour with two secondary cells thus arranged, each of
which gave out, with a platinum wire '2 m.m. in diameter, a light
nearly equal to that of a candle and with a very constant
intensity (71).

M. Trouve recently applied secondary cells for laryngoscopy,
to light the obscure cavities of the human body and obscure
cavities in general. (I) He arranged platinum wires of different
forms, in the focus of little spherical, concave or parabolical
reflectors, ingeniously combined according to the nature of the
cavity he wished to light.

In order to prevent fusion of the platinum wires by too bright
an incandescence under the action of the current from the
secondary cell, M. Trouv4 has added to the apparatus which we
have before described, fig. 25 (109), a platinum wire rheostat
intended to graduate the intensity of the current, according to the
diameter and the length of the platinum wire used as lighting
apparatus, or cauterizer in galvanocaustics.

(i) Bulletin of the meeting of the Societe frangaise de Physique p. 2, Jan. 4th, 1878. —
La Nature, 6th year, p. 107, July i3th, 1878.

95

He has also added a double circuit galvanometer for reading
the charge of the secondary couple, and to discover in what state
the battery for charging it happens to be.

113. Application to firing- mines, etc. — Among the
applications to which the couples of secondary batteries may be
put is the firing of mines, for it requires a calorific effect of short
duration repeated at certain intervals.

We described in i868(I) a secondary battery of small surface
which could be employed for this purpose, when the circuit has
considerable resistance. The battery represented in fig. 20, may
suit still better It is only necessary, in making use of these
batteries, to keep them charged, and not to leave them too long
without working, for they are then more difficult to charge.

A fuse formed by a platinum wire of ?V of a millimetre, dipped
in powder or gun cotton, may be fired by the current from a
battery of 20 couples, with a resistance in the circuit equivalent
to about 6 kilometres of telegraph wire.

If the circuit has a lower resistance, say 300 metres of this
same conductor, one secondary couple may suffice to fire succes-
sively, with a single discharge, a considerable number of fuses.
It is not necessary for these cells to have a large surface. Couples
such as those represented in fig. 26 (no), or even yet smaller,
such as those in fig. 31 (115), may be used and constitute a very
portable apparatus.

114. Fig. 28 represents a portable battery enclosing two small
secondary couples, which connect easily in parallel during the

(i) Annales de Chimie et de Physique, 4th series, vol. XV, p. 26.

OP

T7H 17 BE SIT 7

96

charge, and in tension during the discharge, by means of a com-
bination of three buttons CDC
which serve as a simplified
commutator in this particular
case.

The connections of the
secondary couple are, in fact,
arranged so that, the poles of
the primary battery being put
in connection with the two
edges of the box, if the two
buttons C C are pressed, the
fig. 28, two COUp}es become charged

at the same time, like a single couple of double surface, and if
the button D be pressed after having loosed the buttons C C, and
the connections with the primary battery taken away, the two
secondary couples become combined in tension for the discharge.
These buttons are otherwise arranged like that represented in
fig. 12 (45).

If it is wished to charge each couple separately, independently
of each other, only one button of three should be pressed, either
the left button for charging the couple on the right, or the right
button for charging the couple which is behind the left button.

The three buttons ought never to be pressed all at once, in that
case the apparatus would be short circuited.

These small batteries of two cells may be used in cases where,
the circuit being of too high resistance, ignition would not be
sufficiently instantaneous with a single secondary cell,

97

When simultaneous firing of a great number of fuses is required,
it is merely necessary to multiply the number of the cells or
secondary batteries, without increasing the power of the primary
battery which serves to charge them, and which is always composed
either of two Bunsen elements mounted in the usual way, or, as
we have previously described, three Daniel! elements.

It is true that induction coils have already been used for the
same purpose, but they require the circuit to be more perfectly
insulated, and it is impossible to properly test the circuit with a
galvanometer.

115. Application to domestic uses.— Saturn's "tinder

box." — Figures 29, 30 & 31 represent a novel arrangement, in
which one of the secondary
cells is placed, easily permitt-
ing a light to be obtained in
laboratories and for domestic
purposes. (l)

This apparatus, which we
have designated, according to
the traditions of the ancient
chemists, by the name of p-v

Fig. jo.

(i) Comptes rendus, t. LXXVII, p. 466. i873.-Les Mondes, t. XXXI, p. 747. 1873.

98

Saturn's "tinder box," is composed of a small secondary cell of well
formed lead plates enclosed in a box, on the bottom and sides of
which is arranged a system of connections, so as to enable a
platinum wire to be raised to incandescence, by the simple pressure
of the finger upon a metal tongue, in order to light any inflam-
mable substance, such as the wick of a candle, a mineral oil
lamp, or gas, &c.

The apparatus is charged, and preserves its charge constant by
being placed against two metal plates fixed upon the sides of a
box containing a battery of three Daniell or Callaud cells (32).(I)

This battery may also be placed at a distance, if more con-
venient, and its
poles connect-
ed by wires to
a little board
upon which the
terminal springs
serving as con-
nectors are fix-
ed (33), against
Fig. 32. which are plac-

ed the metal plates of the "tinder box," for charging.

(i) Leclanche cells which are so convenient to use in many cases cannot be advantage-
ously employed in this instance. They would run down too quickly, as the circuit is nearly
always closed in order to keep the apparatus charged. Besides, their E.M.F. diminishes
considerably by prolonged closing of the circuit, so the secondary cells become less highly
charged with three of these elements than with three Daniell or Callaud elements. Callaud
cells are to be preferred, because, the sulphate of copper, remaining in a dense layer at the
bottom of the jars, is less rapidly reduced by the zinc than in the Daniell elements of porous
jars in which the sulphate of copperf filtering through this jar, becomes directly exposed to
the reducing action of the zinc without any advantage to the current.

Three weeks may pass without renewing the sulphate of copper in a Callaud battery used for
maintaining the "tinder box," whereas in the Daniell battery it must be renewed every week.

The Callaud battery only requires more time allowed it to begin working, and to attain
its full E.M.F. the first time it is set up.

99

There is the great advantage in this arrangement of being able
to charge the apparatus, and to remove it charged, without attach-
ing or detaching any connecting wire. Any mistake in the
direction of the primary current is also prevented. The positive
pole of the secondary cell corresponding with the terminal C (fig.
31) must always come into contact with the same terminal' of the
primary battery, whichever way the apparatus happens to be turned.

When the secondary cell has been charged by a prolonged
action of this battery, it is
only necessary to press the
finger upon the metal
tongue arranged to close
the secondary circuit in
order to make it work. The &£• 33-

platinum wire is thus raised to a sufficiently high temperature to
instantly set fire to any combustible substance'1' (fig. 29).

With the store of electricity preserved in this little secondary
cell, after being charged to its full extent by a weak but prolonged
current from the battery, a hundred lights may be consecutively
obtained. It follows that it is not necessary to keep the secondary
cell constantly fully charged by the action of the battery, and the
object of being always in connection is to economise the current
from the battery when it is thought that the secondary cell, not
having been exhausted by a considerable number of successive
discharges, might still produce many lights without being recharged.

(i) It is essential that the wick of the taper or small candle be traversed by the platinum
wire, because if the wick happened to be too far below it would not light so easily, and
besides, if it should catch fire, the platinum wire would be immersed in the hottest part of
the flame, and, being at the same time raised to a white heat by the passage of the current,
it might melt.

100

The lighting of a candle, by means of platinum wire raised to a
white heat, is produced noiselessly and more instantaneously than
by any other means. As the incandescence of platinum wire
does not in any way alter the composition of the air, there is no
smoke or smell of suffocating gas, like that which takes place with
sulphur or chlorates. One need not fear risk of fire, or poisoning
by phosphorous. We may finally take this means of lighting as
very economical; for, on the one hand the secondary cell requires
no expense at all for maintenance, the lead and the solution being
put in once for all without ever requiring renewal, and, on the
other hand, it suffices, in order to keep the weak current of the
charging battery in action, to add from time to time a few crystals
of sulphate of copper, of which the consumption is extremely small
compared with the great number of times the light may be obtained.

The "tinder box," once charged, may be carried away, and, in
consequence of
these cells having
the property of
preserving their
charge, a consid-
erable number of
lights may be ob-
tained without
putting it back in Fig. 34.

connection with the primary battery.

116. Figure 34 represents another novel arrangement of the
same apparatus, making it a kind of electric candle-stick. The
terminals, between which the platinum wire is held, and the little
candle are fixed on a separate board in connection with small
vertical plates of metal.

101

By placing these small plates against corresponding plates upon
the secondary cell, the incandescence of the platinum wire, and
consequent lighting of the candle, is instantaneously produced,
and the latter may be as easily carried away as any ordinary
candle-stick.

117. " Saturn's tinder box " may also be combined with electric
bells, so as to work from one primary battery without in any way
impeding the action of the bells, by placing it in direct communica-
tion with the two poles of the battery, and thus forming a shunt
circuit.

It would appear, during the charge of the secondary cell under

35-

the action of a battery in the circuit of which are placed one or
more bells, that this apparatus must absorb all the current and
prevent the bells working, but as the lead plate secondary cell
acquires under the influence of the current a high intensity, the
result is that it does not act like an inactive derived current but
even contributes towards working the bells. Especially if the
battery itself happens to be too weak to make the bells ring, the
secondary cell becomes capable, by means of the power stored,
of putting them in action. It acts in this case like an accumulator
of work done, a sort of electric fly-wheel.(l)

(i) Comptes rendus, t. LXXVII, p. 466. 1873.

102

Besides, as bells only work in an intermittent manner, allowing
from time to time, sufficient intervals for the secondary cell to be
charged, the cell does not quickly become exhausted, even if the
bells work continuously; for, in consequence of this store of
electricity, a secondary cell well charged will work one or more
electric bells continuously for more than an hour.

The two kinds of apparatus may even work simultaneously
without injuring one another. Thus, at the same time, the candle
may be lighted and the bells sounded. This happens from the
fact that the secondary cell is in the shunt circuit, and as the
platinum wire possesses a considerable resistance, part of the
current passes through the bell circuit.

118. The same apparatus may be applied to lighting gas, and
the more easily because gas does not require the incandescence of
so large a platinum wire as a wax or stearine candle. Consequently
it may be effected at a considerable distance, and if it is a question
of lighting a large number of jets simultaneously, recourse may
be had to batteries composed of a great number of secondary
cells, just as used for firing mines.

119. Application to Electric Breaks for use on
Railways. — We have recommended the use of the secondary
cells above described, wherever a current of great intensity is
required for a very short time, to produce either calorific or
magnetic effects by means of a feeble source of electricity. (l)

M. Achard used them recently with success for working his
electric breaks, which require at a given moment the passage of a
strong current through a series of electro-magnets wound with
thick wire.

(i) Recherches sur Irs courants secondaires et leurs applications, Annales de Chimie
et de Physique, 4e serie, t. XV, p. 20, 1868.

103

The secondary cells are kept charged by a primary battery of
three Daniell elements, the current of which they accumulate,
as in the preceding applications (89, 109, 217).

The accumulated force is then spent, in an instant, in the form
of magnetic work.

The primary battery remains in constant connection with the
secondary cells. Thus, as we have above explained (46), it
combines its feeble action with that of the secondary cells during
the discharge, and again acts upon the secondary cells by charging
them as soon as the circuit of discharge is interrupted.

120. Application to the eudiometrie analysis of the

atmosphere Of mines. — In all cases where a calorific effect
of short duration is wanted these cells may be advantageously
employed. It is thus that M. Coquillon has made use of them
for heating a palladium wire, and determining the combination of
air and proto-carbonised hydrogen in his fire damp detector.(I)

121. Application to the production of Luminous
Signals. — The secondary batteries described (96), are able to
produce a voltaic arc of several seconds duration and of very
great intensity by employing a sufficient number of secondary
cells, charged for a few minutes from two Bunsen elements, and we
have noted the use that might be made of them in certain
instances for the production of luminous signals. (2)

Although this idea has not yet been put into practice, we feel
sure that it would be of great service at sea, or along the coasts,
because the inconvenience and cost resulting from the use of
batteries become considerably reduced when it is only a

(i) Comptes rendus, t. LXXXV, p. 1106, 1877.
(a) Brevet du 29 F^vrier, 1868.

104

question of setting up two Bunsen elements in order to obtain at
any moment an electric light equal to that which eighty or a
hundred of these elements will give direct.

Mons. A. Niaudet(I) has taken a special interest in this
application, and recommended the use of the Gramme machine'2'
for charging secondary batteries intended for the production of
luminous signals, by means of mechanical force, which signals
would often prevent collisions at sea.<3)

Mons. J. Morin(4) has made several experiments with the same
object by means of a secondary battery of fifty large cells worked
by a small magneto electric-machine, having only eight coils.
This battery could fuse, in discharging, an iron wire 2-20 metres in
length and o'ooi metres in diameter. Mons. Morin has been
occupied in the construction of a special electric lamp for
producing the voltaic arc under these conditions.

122. Application to the production of the Electric

Light in special cases.— In certain cases, where a bright light
is required for several minutes only — in projection experiments for
example, or for any other study — the battery of six large cells
above described (100) answers the purpose.

Mons. Reynier's electric lamp allows of obtaining, some good
results, even with this low tension.

As this lamp possesses considerable resistance if carbons of
small diameter are used, the expenditure of the electricity stored

(1) V. La Nature, 27 Juin, 1874.

(2) The Gramme Machine has been used for several years in the workshops of
Mons. Breguet for forming and putting secondary batteries into working order.

(3) It is known that Mons. Treve has studied in a special manner, and by other means,
the solution of this important question.

(4) Comptes rendus, t. LXXXI, p. 435, 1875.

105

during two hours in the secondary battery is comparatively slow,
and, consequently, the light may last for a quarter of an hour.

By making use of a larger number of these big cells, connected
together in two or three batteries, a very bright light could be
obtained, long enough to be of use in certain cases.

123. Application to the sub-division of the Electric

Light. — The lead plate secondary battery composed of forty
small elements, designed to produce continuous effects, of which
we have previously spoken (101), could, by giving a sufficient
rapidity to the commutator, supply a continuous voltaic arc, of
proportionately reduced power it is true, but obtained by means of
only two ordinary Bunsen cells. Two large size cells would keep
several similar arcs going by working through a considerable
number of similar small batteries, and there would be thus
obtained a solution to the problem of sub-division of the
electric light, by means of secondary currents.

This solution, which we pointed out some twelve years
ago,(l) appears rather complicated. Still, it is not impossible to
carry out; for commutators constructed like those described
further on (part V.), do not require a high E.M.F. in order to keep
them going; they could be arranged so as to turn altogether like
the bobbins of spinning looms, and besides, secondary batteries
employed for this transformation could be put into quite a small
space, as they need have but a very reduced surface.

Mons. W. Lermantoff has devoted himself to experimental
work upon this subject.'2'

(1) Brevet du, 27 Avril, 1868.

(2) Journal de Physique, t. 258, 1876.

106

124. Physiological Effects produced by Secondary
Batteries. — The E.M.F. of each secondary element of lead
plates being, as already noticed (80), fairly strong, secondary
batteries of twenty or forty cells suffice to give very powerful
physiological effects. These effects could be utilised in
therapeutics, by using secondary cells of very small surface, so as
to avoid any effects of temperature. The secondary battery for
continuous currents, of which we have already spoken (101),
would be suitable for this purpose; even batteries intended for
only temporary work could be employed, as the discharge of these
would last long enough to produce an effective action, in
consequence of the high resistance of the human body.

125. Various Applications. — Secondary cells and batteries
as above described, may be, generally, applied in every instance
where, at a given moment, a powerful electric effect, either of
quantity or tension, is required.

Such is the case, for example, when it is a question of
transmitting the time of day simultaneously to several different
places by means of a current through a number of wires.

These apparatus may be of great use in scientific research,
as we shall see further on (Part III).

Mons. Thore, of Pau, in 1875, employed a light furnished by a
small battery of twenty cells for spectroscopic experiments.

Mons. Guerin in 1875, used these cells in electro-chemical
gilding and silver plating, in certain cases where a very strong
current, of quantity, was required for a short time.

126. Other applications are recommended, based upon the
results of our research in voltameters.

107

Firstly, in 1860, the substitution of lead electrodes for those of
platinum, used by Jacobi to produce counter polarisation currents,
for preventing delays in the signals upon certain telegraphic lines
which were imperfectly insulated.

Secondly, in 1865, the use of lead anodes instead of platinum
for electro-typing.

127. A fact to which we have drawn attention when specially
studying copper wire voltameters ;(l) viz. the formation of a very
fine point at the extremity of the positive electrode, has been
the subject of some experiments by the engineer, Cauderay, of
Lausanne, for the pointing of pins by electro-chemical means.

Apparatus made upon this principal was exhibited at the
International Exhibition of 1867, and, if this application has not
been followed up since the death of Cauderay, it no less deserves
to be noted and again taken up, on account of the unhealthiness
resulting from the mechanical process of pointing pins, &c.

128. The phenomenon we noticed in i859,(2) and which we
have previously described (23), viz. the almost complete cessation
of a current traversing a voltameter of alluminium electrodes, by
reason of the insolubility of the alluminium formed upon the
positive pole, has recently furnished Mons. Ducretet with an idea
for some ingenious appliances in telegraphy. (3)

(1) Bibl. univ. de Geneve, t. VII, p. 332, Avril, 1860.

(2) V. Comptes rendus, t. LIX, p. 610, 1859.

(3) V. Bulletin des seances de la Societe francaise de Physique, p. 17 et 40,
Janvier-avril 1875.

PART.

Effects produced by Electric Currents
of high Tension.

CHAPTER. I.

Luminous Sheath. — Luminous Liquid Globules. —
Globular Flames. — Voltaic Brush. — Luminous
Figures. — Moving Spark. — Bunch of Aqueous
Globules. — -Jets of Steam. — Electrified Liquid
Vein with Spiral Motion. — Electric Bar. —
Voltaic Pump. — Electro-Silicious Light. — Crowns,
Arcs, Rays, and Undulating Movements.—
Electro Dynamic Spirals. — Crater-like Perfor-
ations.

129. The secondary cells above described (96 — 98) have
permitted of the study of phenomena produced by electric
currents of high tension, and particularly those which appear in
the passage of these currents through liquids.'11

(i) Research in Phenomena produced by Electric currents of high tension in liquids.
Comptes rendus, t. LXXX, p. 1133, 5 Mai, 1875.

109

Some phenomena of this kind have already been made the
subject of study with ordinary batteries by Davy, Hare, Makrell,
Grove, Gassiot, de la Rive, Wartmann, Despretz, Fizeau & Foucault,
Quet, Maas, Van der Willigen, &c ; but the necessity of having
to set up a powerful battery for their observation was such a
difficulty that they have never been deeply analysed. The currents
supplied by secondary batteries are, it is true, but temporary; they
have, nevertheless, sufficient duration to enable us to follow in all
their details the effects produced by the passage of electricity
through imperfectly conducting substances, such as the solution in
voltameters ; besides, the experiments may be renewed by recharg-
ing the apparatus, and the intensity of the current slowly diminishing
as the discharge continues, far from proving inconvenient, success-
ively shews to the observer a series of different phases, which would
escape remark with a constant current, or would require continual
modifications in the elements of the battery.

The study of these phenomena also presents so much the more
interest, because, at this point are found the two powers which
exercise the most direct influence over the elements, viz : electrical
force and chemical force, where a solution for every problem
in human industry appears to be found.(l)

In fact, by following the passage of currents of varying intensity
through liquids, we watch, so to speak, the contest between "the
electric flow" and molecular attraction, added to chemical affinity,
tending to hold united the metal molecules of the electrodes, or
the elements of the liquid body contained in the voltameter. If
"the electric flow" possesses great intensity, mechanical and

(i) Dumas. Bulletin de la Societe d 'Encouragement, t. XII, p. 153, 1866.

110

calorific effects predominate ; the molecular attraction is first over-
come, and the electrodes are disintegrated, fused, or vapourised.
If the intensity is somewhat less, the electrodes become the seat of
luminous phenomena, produced by the vacuum and rarified
vapours around; the solution, hardly moistening the electrodes, is
scarcely decomposed. If the intensity of the current still decreases,
most of the calorific and luminous phenomena disappear, and
chemical decomposition is exhibited; and as, on the other hand,
the current then passes more thoroughly in the liquid, its intensity
seems greater throughout the circuit. This may be shown in a
very striking manner by the following experiment.

130. Experiment upon the "luminous sheath" with
the current decreasing in intensity.— The current from the
secondary batteries, each composed of twenty pairs of lead plates,
is discharged through a voltameter V, of water acidulated by
sulphuric acid, and platinum wire (36). The positive wire only is

Fig. 36.

first dipped in the liquid. A galvanometer G, is already
arranged in the circuit and a platinum wire F, stretched through
the open air, about 80 centimetres long and o'i m.m. in diameter,
is also placed in the circuit. Immediately the negative platinum
wire is dipped in the solution there is seen upon this wire, without
any noticeable liberation of gas, a luminous sheath such as has
been observed with ordinary batteries by the above-mentioned
scientists. The positive wire, on the other hand, sets free but a

lit

very small quantity of gas. The galvanometer only shews a slight
deviation, and the platinum wire stretched through the air is not
reddened. But if we leave the experiment alone, in two or three
minutes, as the E.M.F. of the secondary battery falls, the luminous
sheath disappears, and an abundant liberation of gas takes place
suddenly at both poles ; the galvanometer shews a strong deviation,
and the platinum wire is at the same moment rendered incandes-
cent throughout its length.

131. The various phenomena which may be produced with
different metals and solutions, according to which pole is first
dipped, and which have been studied with great care by Van der
Willigan,'1' by means of a Bunsen battery of forty elements, are
easily re-produced with a secondary battery of forty cells, and we
believe that the following rule holds good for these phenomena
under the conditions in question, viz: the electrode first dipped in
the solution, or the one which presents the larger surface sub-
merged, gives its sign to the liquid in the voltameter.

132. Change of colour in the "luminous sheath"
according to the intensity of the current.— In proportion

as the battery is discharged and the intensity of the current
decreases, we have noticed that the colour of the luminous sheath
formed round the negative electrode gradually alters; it passes
successively from white to blue and violet, and towards the last,
a few seconds before the liberation of gas appears, it is reduced to
a few brilliant points of a reddish purple, which surround the end
of the electrode.

At first we thought there might be some possible connection
between the intensity of the electricity in play and the refrangibility

(i) Annales de Poggendorff, t. XCIII, p. 285.

112

of the light produced; but later experiments with currents of
greater intensity having more clearly shewn the nature of the
phenomena which take place around the electrodes, these changes
of colour may be explained in the following manner : — The lumin-
ous sheath is nothing else than an envelope of rarified and
incandescent gases formed round the electrode, and of equally
rarified and incandescent vapour from the solution of the
voltameter.

What is the nature of these gases ? In consequence of the
very high temperature produced round the electrode with a current
of great intensity, the water is partially decomposed around the
-same pole, as proved by Grove, and as our own researches have
verified. There is then hydrogen, oxygen, and sulphuric acid, or
sulphurous vapour, round the electrode, when the liquid is water
acidulated by this acid. It can also be easily understood that
the nitre arising from the atmosphere may be held in solution by
the liquid. All these elements are rarified and luminous, and the
colour of the light necessarily participates in the mixture. A white
tint predominates, probably arising from the relative abundance of
sulphurous vapour given out by the liquid. If salt water is used
the sheath emits a brilliant yellow light, due to the excess of
sodium.

But, as the intensity of the current decreases and the heat dim-
inishes, these dissociations become less complete, the proportions
of the various products become modified, and consequently the
colour varies. (I) When the current is reduced to a very feeble

(i) We also notice in Geissler tubes submitted to a prolonged action of an induction or
high tension current, similar changes in the colour of the light emitted ; but in that case the
alterations are not caused by variations in the intensity of the current ; they are permanent
and are due to modification in the gaseous matters contained in the tubes, the limited
quantity of which is not renewed.

113

intensity, the calorific action lessens and the point is approached
where electrolysis of the water is produced in the ordinary manner ;
hydrogen begins to dominate alone at the negative pole, and, if
the heat caused by the current is still sufficient, the latter period
of incandescence is maintained for a few moments.

Hence this purple colour in the light, which finally appears at
the extremity of the negative electrode; for we know that such is
the colour proper to incandescent hydrogen enclosed in a narrow
space. Now, the luminous sheath here becomes so much the
more compressed by the vicinity of the liquid, which the calorific
effect of the current has further diminished.

133. Batteries of from two hundred to eight hundred
secondary cells, used in studying" electrical effects of

high tension. — In order to observe effects produced by electric
currents of very high tension, we successively connected up
batteries of two hundred to eight hundred secondary cells, the
discharge of which, for the first few moments after the action of
the primary current, equals that of from three hundred to twelve
hundred Grove or Bunsen elements.

The E.M.F. of each secondary cell, immediately upon breaking
the primary current, is, we know, equal to one and a half Grove or
Bunsen elements, as above shewn (80). This E.M.F., it is true,
undergoes some fall when the secondary circuit is not immediately
closed upon the primary circuit being broken ; but, in spite of this
fall, it still remains higher than either a Grove or Bunsen cell (81).

The resistance of the cells composing these batteries is, on the
other hand decidedly lower than that of the Bunsen cell of
ordinary dimensions, in consequence of the very close proximity
of the lead plates, and in spite of their limited surface (two square

114

decimetres). This resistance is hardly equal to three metres of
copper wire one m.m. thick (83).

The result is that each of these small secondary cells is capable
of producing, when well charged, a calorific effect sufficient to
raise a platinum wire, '3 to '4 m.m. in diameter by five cents, in

y-j-^^H t*>

length, to a red heat. With t wo „ secondary cells only we have
been able to redden a platinum wire of this diameter and ten
metres long.

This incandescence is certainly of very short duration in
consequence of the small surface of each plate ; but if the
discharge take place through circuits of higher resistance, for
instance, if we study its effects uppn the surface of a liquid, the
expenditure of current is much less rapid, and, with a single
discharge, we have often been able to repeat more than twenty
experiments without completely exhausting the charge in the
battery.

134. Figure 37 represents an arrangement of four hundred
secondary cells divided into ten batteries, each of forty pairs.
These batteries are of the same shape as those already described
in figure 20 (98) ; but they are composed of double the number
of cells.

In our later experiments, made with eight hundred secondary
cells, a second series of batteries exactly similar was arranged in
another room, and the current connected by wires to that of the
first series.(I)

These batteries, coupled up first in parallel by means of
commutators, only require two or four Bunsen cells in order to be

(i) Still more recently we arranged these batteries in steps upon shelves, so as to
allow of their being placed in a still smaller space.

115

116

charged all at once, and these Bunsen cells may be placed
outside a window, upon the sill, so as to avoid the acid fumes.
When the batteries have not remained too long without working, a
few hours are sufficient to charge them.

Then, by turning the commutators, all the secondary cells may
be connected in series, and the charge resulting from the chemical
work accumulated during several hours by the two or four Bunsen
cells may be expended at will, either in a few seconds, or in any
longer time.

Experiments are most often made in the dark, so that the
details of the luminous phenomena produced may be studied.
The voltameter is shewn at the moment when the electric current
happens to act upon its surface. The water vapour may still be
seen to be set free above the liquid, after the powerful calorific
effect produced by the passage of the current.

Some rheoscopes of platinum wire, the same as that shewn in
figure 24 (102), are placed upon the tables, and serve to test the
condition of the secondary cells, in which some accident may
occur, as above explained.

Other large rheoscopes, with a long platinum wire stretched
between the terminals, allow of the separate examination, if
necessary, of the condition of each battery. (l)

(i) When it is required to put all the batteries into work, the experiments are rather
delicate and need care in preparation, on account of the multiplicity of secondary cells and
great number of metallic connections necessary.

They are also not free from danger in carrying out, for the discharge of such currents
combining, at one and the same time great quantity and potential, can produce very violent
shocks upon the human frame. For three years we were fortunate enough to avoid any-
thing of this kind ; but upon the occasion of an experiment made lately, in order to charge
the rheostatic machine, which will be described further on (part V.), having involuntarily
touched the naked ends of the wires from a series of six hundred secondary cells, we
instantly felt, not only a very strong shock, but the sensation of burning throughout the
body, rising as far as the neck, causing us to cry out, and considerably frightening the
people standing round. All the same, this accident had no unpleasant consequences. But
it might not perhaps have been so if the eight hundred secondary cells had then been in
action. Shocks given by induction coils never seem to produce the same effect.

117

135. Luminous Liquid Globules.— If a battery of two
hundred cells be put in connection with a voltameter of water
acidulated by sulphuric acid, or of salt water, so that the positive
wire alone is immersed to begin with, the approach of the negative
wire towards the liquid causes the fusion of this wire, or its
vaporization, with a kind of explosion, and a flame variously
coloured, according to the kind of metal forming the electrode.

By diminishing the portion of acid
contained in the solution of the
voltameter, so as to avoid complete
fusion of the metal, a continuous series
of sparks, accompanied by a cracking
noise, is produced, and these sparks
continue for several minutes, decreasing
gradually in intensity (Fig. 38).

g- 38-

But if, when the negative wire is
immersed first in the solution (which
ought to be by preference salt water

Fig. 39. Fig. 40.

so as to avoid acid vapours, and slightly raise the resistance

118

of the circuit), we bring the positive wire close to the surface
the result is quite different.'1'

The wire is not melted and a small luminous liquid globule,
accompanied by a curious noise, is seen formed at its extremity
(Fig. 39). On gradually withdrawing the wire the globule
increases in size, as if the liquid was sucked up by the electrode ;
it acquires a diameter of about one centimetre and takes at the
same time a rapid spiral motion. In consequence of this motion
it becomes flattened (Fig. 40), — elongated sometimes towards the
negative wire, if it is near enough, — and is finally dispersed at the
same time that a detonating spark is produced at the negative pole
when this pole only dips a very little way in the liquid. The
globule is again spontaneously formed at the end of the positive
wire, and the same phenomena thus take place several times in
succession in an intermittant manner.

136. The spiral movement does not always take place in the
same direction, like the magneto-electric spiral motions described
further on (158). It takes place sometimes in one direction and
sometimes in the other. It often happens to go in the same
direction a number of times running; but this may change
without any apparent cause. It is a re-action spiral motion,
similar to that of electric whirls, and is due to the flow of the
electric current through the liquid.

The globule becoming nearly detached by its spherical form
from the rest of the liquid, or only having a very small surface in
contact with this liquid, the motion takes place in either direction,

(i) The voltameter is placed upon a support furnished with crooks, to which are fastened
platinum wires in connection with the poles of the battery, so that they may be introduced
with proper care into the liquid.

119 ,

according to the position of the point upon the surface of the
globule through which the main flow of current passes or the
liberation of vapour is produced

The luminous appearance of the whole of the globule seems to
arise from the bright light emitted at the point of contact with the
rest of the solution.

The noise is due to condensation of the steam, which tends to
form round the electrode, in the liquid. The intermittant spark,
which appears at the negative pole at the instant the globule
attains its maximum development, is explained by reason that the
negative wire, at first dipped a little way into the liquid, soon
becomes separated from its surface by the vaporization of the
portion of liquid which forms the globule. The current is then
broken for a moment, the liquid from the globule, falling back
into the voltameter, re-establishes the contact, and the phenomena
may be thus re-produced several consecutive times, spontaneously,
until the secondary cells are exhausted.

137. As to the concentration of the liquid in this globular
form, it would seem to be explained by the phenomenon of
"suction," resulting from the flow of the electric current
itself at the positive pole. We shall see further on a still
more striking example of this "suction," when a current of
higher tension is used, and when the space for liquid round the
electrode is limited by being enclosed in a narrow tube (148)
("voltaic pump"). But, in this case, the liquid, having unlimited
space, naturally concentrates into the form possessing the smallest
possible surface, consequently becoming spheroidal.'1'

(i) This spheroidal form taken by a liquid, under the action of the calorific effect
produced by an electric current, may be compared with that likewise shewn by liquids
under the action of heat alone, when placed upon red hot surfaces, which have been studied
by Boutigny. It is also the form taken by liquids when simply withdrawn by the action
of gravity, as shewn by the experiments of Plateau.

120

138. Globular discharge.— Brush discharge and lum-
inous figures produced by the discharge of a battery
Of 800 Secondary Couples. — For studying effects produced
with a voltameter of distilled water, we have quadrupled the
potential of the current, by uniting 20 batteries, each composed
of 40 couples, forming a grand total of 800 secondary cells.'1'

When the current from the whole of this battery is passed
through distilled water, there is exhibited at first, in a more intense
degree, a phenomenon somewhat similar to that observed by
Grove, with 500 elements of his nitric acid battery. The
positive electrode being previously immersed in the distilled water,
by bringing the negative platinum wire close to the surface of the
water, and raising it immediately, there is obtained a yellow flame,
nearly spherical, about 2 centimetres in diameter (fig. 41). The

platinum wire, from one to two
millimetres in diameter, melts rapidly
and keeps for several moments in a
state of fusion, at a height of from 14
to 15 millimetres above the liquid.

This flame is caused by rarefied
incandescent air, by the vapourised
metal of the electrode, and by the
elements of the decomposed water;
spectral analysis clearly shows the

presence of hydrogen.
Fig. 41.

If, in order to prevent fusion of the medal, the intensity of the
current be diminished by interposing a column of water in the
circuit, the spark appears in the very clearly defined form of a

(i) Comptes rendus, t. LXXXV, p. 619, October, 1877.

121

little globe of fire, from 8 to 10 millimetres in diameter (fig. 42).

By slightly raising the electrode, this globe takes an ovoid form;
blue luminous points, their number continually varying and
arranged in concentric circles, appear on the surface of the water
(fig. 43). Rays of the same colour soon shoot forth from the
centre and join these points (fig. 44).

From time to time, the rays take a gyratory motion, sometimes
in one direction, sometimes in another, describ-
ing spiral curves (fig. 45 and 46). Sometimes
the points and rays all disappear from one side,
and varied curves, formed by the movement of
those which remain, become visible on the sur-
face of the liquid. Finally, when the rapidity
of the gyratory movement increases, all the rays
vanish, and nothing more is seen but the blue Fig. 42.
concentric rings (fig. 47). The rings are the last phase of these
transformations, which are very curious when followed with the

Fig. 43. Fig. 44. Fig. 45. Fig. 46. Fig. 47.^

naked eye or with a glass, and constitute quite an electric
kaleidoscope. (l)

(i) These phenomena may be compared with those observed by M. Fernet with induction
currents; Comptes Rendus, 1864; they also greatly resemble those resulting from the fall of
liquid drops on a level surface, studied by MM. Helmholtz, Thomson, Maxwell, Tait,
Rogers, Worthington and Trowbridge.

122

139. The production of these figures is explained by the
excessive mobility of the arcs or luminous fibres which compose
the ovoid light, between the water and the electrode. By care-
fully examining this particular form of spark, it is discovered, that
in reality it is a kind of tuft or voltaic spray, similar to that in
static electricity, but more complete, by reason of the greater
quantity of electricity in play. These luminous fibres, being in a
continual state of agitation, the points where they meet the surface
of the liquid constantly change, and form the rays which are seen.
Their gyratory motion arises from the reaction due to the electric
flow. As to the rings, they visibly form under the eye of the
observer, by the motion of the blue points becoming more and more
rapid, and by the continuance of the impression upon the retina.

140. When the metal electrode is positive and the distilled
water negative, the spark again takes outwardly an ovoid form,
but the middle is traversed by a cone of voilet coloured light.

When two metal electrodes are used, a luminous spheroid is
obtained, the interior of which is traversed by a brilliant streak of
light. This appearance corresponds with the streak and the halo
of the spark from induction currents; only here, the halo occupies
more space, in consequence of the greater quantity of electricity.
In short, if the length of the column of water interposed be much
increased only an arc or rectilineal ray of light is obtained.

It is not necessary, in these experiments, to bring the electrode
into contact with the water, to cause the passage of the electric
current. The tension of the batteries, although the couples which
compose them are not insulated in any special way, is great
enough to make the spark strike spontaneously at the distance of
about one millimetre above the liquid.

123

141. If, instead of leaving the electrode fixed on the surface
of the voltameter, during the flow of the electric current in the
form of these sparks or globular brushes, one of the wires serving
as electrode be suspended at a considerable height, and enough
weight and length given to it to cause it to oscillate like a
pendulum on the surface of the liquid, or over a conducting plate,
without perceptibly changing its distance from this surface, the
little globule of fire, produced at the extremity of the wire,
naturally follows the movement of the electrode, and, when working
in the dark, nothing but the globule of fire is seen to move on the
surface of the liquid. Further on, we shall refer to this experiment
(part IV.) in order to explain certain appearances in natural
electric phenomena.

142. Wandering electric sparks.— The electric spark in
this globular form, resulting from the action of a great quantity of
electricity upon ponderable matter, may be excited to a progressive
motion, by itself, without it being necessary to move either electrode.

This is the result of a more recent experiment we have made(l)
by using the apparatus described in the Fifth Part under the name
of Rheostatic Machine.

Although this experiment does not necessitate the use of a
voltameter, we will here give a description of it, because it relates
to the globular forms of electrified matter of which we have just
quoted several examples, and we will compare this experiment with
the preceding ones, to explain, by analogy, the slow progression, in
certain cases, of globular lightning.

If the two poles of the secondary battery of eight hundred
couples are put in connection with the armatures of a condenser,

(i) Comptes rendus, vol. LXXXVII, p. 325, August igth, 1878.

124

the insulating plate of which is formed by a sheet of mica, this
condenser becomes charged like a Leyden jar, and can give, when
discharged, a spark of the nature of static electricity.

But if, by chance, the mica plate has some very thin place or
crack, made when it was cut out, it is spontaneously pierced
at this point by the action of the current from the eight
hundred secondary cells, the same as the glass of a Leyden jar
over charged by an electric machine.

A remarkable phenomenon is then presented to view. In
consequence of the great calorific power of the electricity in play
in this experiment, the spark which has struck across the condenser,
between the two armatures, is not instantaneous like that of static
electricity, but, as it is accompanied by fusion of the metal and
even of the insulating material of the condenser, it forms a small
and very brilliant luminous globule which moves with a peculiar
noise, and slowly traces, on the tinfoil of the condenser, a deep
track, sinuous and irregular.

Figure 48 represents a true copy of that portion of the surface
of a condenser where the phenomenon occurs. The spark appears
at A, branches out soon at B, as far as C ; there it disappears, to
reappear immediately at the point B, with such rapidity and in an
interval of time so hardly appreciable, that it seems to have made
a bound ; next, it goes towards D ; there it forms a new branch
which stops at E. reappears at D, continues its way towards F, and
so on. Sometimes, as in the present case, the spark appears again
farther on, at the point Q, detached from the principal track, to
stop next at R, and the phenomenon only ceases when the mica
plate presents no other part thin enough to be pierced. In other
cases the spark remains some time stationary round the same

125

point, other times again, one of the branches is immoderately
prolonged, and describes, on the entire surface, outlines similar to
those on a geographical map. A tube of distilled water is

/%-. 48.

previously interposed in the circuit of the secondary battery, to
prevent too intense calorific effects and deflagration of the whole
condenser.

Whilst the phenomenon is happening, the points through which
the spark will pass cannot be foreseen ; nothing is more curious
than the motion of this little shining globule, which is seen slowly

126

making its way, choosing the points towards which it must go,
according to the greater or less resistance of these points in the
insulating plate.

The condenser is perforated by the passage of the spark, and
the tinfoil forms a double row of melted beads around the edges
of the consumed mica.

Fig. 49.

143. Sheath Of aqueous globules.— Taking again the
solution of salt voltameter in which the current from a battery of
two hundred cells produces at the positive pole a luminous liquid
globule; if the tension of the current happens to be doubled, by
employing a battery of four hundred cells, the effects are
completely changed.

There is then obtained, by the immersion of the positive wire,
instead of a single globule, a sheaf of innumerable ovoid globules
which succeed each other with excessive rapidity, and are
projected to the distance of more than a metre from the jar in
which the experiment is made (fig. 49). It is a kind of
pulverization of the water into small drops, produced by the
electric discharge.

127

The spark in this case appears at the surface of the liquid, in
the shape of a crown or many pointed halo, from which spring out
aqueous globules.'1' The metallic form of electrode is not
necessary to obtain this effect ; a fragment of filter paper,

moistened with a solution of salt connected with the positive pole,
also produces this phenomenon (fig. 50).

144. Jets Of Vapour. — If the current, instead of encoun-
tering a deep layer of liquid, finds only a moistened surface, such
as the sides or inclined bottom of the jar, the calorific effects
predominate, the halo is more brilliant and the water is quickly
vaporized.

The action of the current differs then according to the
resistance opposed to it, and here is found another example of
reciprocal substitution of the heat and mechanical work resulting
from the electric shock. When the work, represented by the

(i) Comptes rendus, vol. LXXXII, p. 314, January 3151, 1876.

128

violent projection of the liquid, appears, there is neither heat nor
vapour developed, and when no visible work is accomplished,

when the liquid is not thrown about, heat is generated and vapour
escapes.

145. The formation of these luminous tracks, accompanied by
jets of vapour, is intermittent. Each time, in fact, that the
electrode in contact with the moistened surface has vaporized the
small drops of water which surround it, the current is for an
instant interrupted, but another portion of the liquid mass which
moistened this surface flows in immediately, and the phenomenon
begins again, thus appearing intermittently, until the voltaic
discharge be exhausted.

146. Electrified liquid vein— gyratory motion.— If a

vein of saline water is made to flow from a funnel in connection
with the positive pole of the same battery (400 secondary couples),
into a basin in which the negative wire is already immersed,
and beneath which is placed an electro-magnet (fig. 52), there
may be perceived, after closing the voltaic circuit, a luminous

129

thread, accompanied by some brilliant points, in the lower part
of the vein; crackling sparks spring forth at its extremity, vapour
escapes from the water, and the liquid which surrounds the lower
part of the vein makes a gyratory movement in the opposite
direction to that made by the hands of a watch, if the pole of the
electro-magnet be north, and tn the same direction, if that pole
be south. The movement is rendered visible by light substances
spread on the surface of the liquid. (l)

Fig. 52.

If the vein be shortened so as to prevent any break in its lower
part, the electric luminous signs disappear almost completely; the
liquid nevertheless becomes heated, as shown by a slight vapour,
and the gyratory movement becomes more clearly defined and
more rapid. By again prolonging the vein, the luminous electric
manifestations re-appear as before.

(i) Comptes rendus, vol. LXXXII, p. 220, January i7th, 1876.

130

147. Electric Bar. — By placing the positive electrode against
the sides of a jar of saline water in connection with the negative
pole, there may be observed, besides the luminous tracks and
abundant jets of vapour, a violent eddy in the liquid, forming a

Fig. 53-

kind of electric bar^ which raises the water to the height of
\\ centimetres above its natural level (fig. 53). If the currents
meet at certain points with inequalities of resistance, it may
divide and give rise to two or three aqueous hillocks, as
indicated in figure 54.

Fig* 54-

This phenomenon is another result of calorific effect, produced
by the current upon the moistened surface which it encounters.

131

The liquid is repelled by the pressure of the vapour suddenly
developed by the current at a fixed point.

This effect may be compared with the breath or draught produced
by an abundant flow of static electricity. Only, in the latter case,
if the tension be greater, the quantity of
electricity is much less ; also, if a current of
static electricity produces an effect of this
kind only upon air, it could not act in the
same way upon a mass of liquid.

148. Voltaic Pump. — Remarkable
effects of suction may also be produced by
the electric current. If the positive wire be
placed in a capillary tube, leaving however
about half a centimetre free at its extremity,
immediately the electrode tube is immersed
in the salt water, the liquid is seen to rise
with extreme rapidity, to a height of from 25
to 30 centimetres and fall again in a sheet of
brilliant sparks and jets of vapour (fig. 55.)

A voltaic pump is thus formed, in which
the vacuum results from the production and
the condensation of the vapour around the
Fig- 55- electrode.

149. The luminous ring which accompanies the fall of the
liquid from the upper to the lower part of the tube, and reappears
again spontaneously and intermittently in the upper part, only to
again descend, constitutes one of the most brilliant and curious
effects that we have observed with electric currents of high tension.

132

This phenomenon is explained as follows : the liquid sucked up
constitutes a prolongation of the negative electrode itself, formed
by the liquid of the voltameter (131). Calorific and luminous
phenomena should then be seen at the extremity of the liquid
sheet, falling again from the top of the tube, especially if the glass,
already moist, offers a certain conducting power, thus establishing
an outer connection with the voltameter. The intermittent action
arises from the quantity of liquid sucked up being in all very
small, its flow along the sides of the tube does not begin until the
drop formed at the upper part has acquired a certain volume, and
the liquid flows in less time than the drop takes to form. A part,
besides, is vaporized in proportion to the calorific action of the
current.

150. The ascent of the liquid is so rapid, in spite of the
resistance presented by the narrowness of the passage, that the
small luminous drop may be seen at the upper end of the tube as
soon as the lower part touches the liquid.

If the tube be too long for the solution to reach the upper
part and fall again outside, the liquid remains at a certain height,
which gradually becomes lower as the current of discharge
from the batteries becomes weaker. The height of this
column varies according to the tension of the current so that it
could be used as a gauge of the E.M.F.

The tension of a battery has often been compared with the
greater or less height of a column of water, and the quantity of
electricity that it supplies with the more or less lavish flow of the
liquid, according to the diameter of the outlet. The resemblance
is here in a manner materially realised by the mechanical effect of
the electric current.

133

151. Liquid cones.— The effects of suction produced by
the electric current may also appear in another form. If a higher
tension be used, — that of 800 secondary cells, — and if the electrode
be brought near to the surface of the distilled water, the liquid
sometimes rises in the form of a cone before the spark strikes.
This phenomenon, little noticed in static electricity, but neverthe-
less discovered by Peltier, and which has been for a long time
better seen on oil than on water, is much more marked with this
kind of electricity, by reason of the greater quantity in play.

If the electrode be formed of a moist pencil of asbestos or filter
paper, there appears round the little aqueous cone, which remains
suspended at the end of the electrode during the passage of the
current, a thick crown of vapour, proceeding from the
calorific effect developed by the current in its passage through
the liquid.

152. Detonations produced at the ex-
tremity of the positive electrode.— If the

positive platinum wire be placed in a capillary
tube, as in the preceding experiment, but so that
it terminates at the extremity of the tube, when it
is deeply plunged in the liquid, a shrill noise is
heard, and if it be raised after having been two or
three seconds immersed, when the end of the tube
attains the upper level of the liquid, a detonation
is heard similar to that of a fulminating cap. The
tube is nevertheless neither broken nor cracked ;
but the lower opening has become conical and
the glass hollowed in the form of a funnel.(l)

Fig. 56 Fig.

(i) Comptes rendus, vol. LXXXI, p. 185, July 26th, 1875.

134

The intensity of this noise is remarkable when the narrowness
of the annular space comprised between the platinum wire and the
sides of the capillary tube is considered, and if it be also taken into
account that this tube be open at both ends ; however this pheno-
menon is more easily produced if the tube be closed at the top.
Figure 56 represents the tube before the experiment, and figure
57 the same tube after the detonation has been heard.

153. This phenomenon does not happen, or is much less
evident, with the negative wire ; for, in this case, all calorific effect
of the discharge is carried by the electrode, melts the metal and
volatilizes it; whilst, if the wire brought close to the liquid be
positive, it reddens without melting, and it is upon the liquid,
which itself represents the negative pole (131), that the calorific
action of the discharge takes place ; vaporization takes place with
extraordinary energy round the electrode compressed inside the
glass tube, ajhd the shrill noise above mentioned results.

154. We thought at first that the detonation itself, accompanied
by pulverization of the glass, was an effect owing to the sudden
entrance of air in the tube at the moment when vaporization
of the liquid ceased by the interruption of the current. But
having since observed that some bubbles of gas appeared at the
end of the positive electrode, amidst the liquid vortex caused by
condensation of the vapour, and that this gas was formed of an
explosive mixture of the elements of the water, consequent upon
decomposition at the high temperature produced, and the detona-
tion heard at the moment when the liquid leaves the tube, appears
to us to be explained by the ignition of the detonating mixture,
which fills the lower end of the capillary tube under the influence
of the spark.

135 *

This phenomenon would then be allied to those of the same
kind observed in voltameters by M. Bertin with much less tension,
when the gases produced by the two poles become mixed in one
receiver above the electrodes.

There would however be this difference ; that the phenomenon
here occurs at only one pole in consequence of the extremely high
tension of the current.

It is essential, in fact, that the current should have an extremely
high tension for the experiment of which we speak to succeed ;
for two hundred secondary couples are not sufficient ; from two to
three hundred must be used.

To explain the pulverization of the lower part of the glass tube,
an effect that the explosion of a mixture of very small quantities
of oxygen and hydrogen barely accounts for, one would be inclined
to suspect the formation of some very unstable compound of
chlorine which is known to occasion combinations gifted with
great explosive power ; the more so as we have not been able to
observe the effect in question with other liquid than solution of
chloride of sodium.

Whatever may be the cause of the phenomenon, it has appeared
to us worthy of remark, and we purpose some time to make a
special study of it.

155. Eleetro-silicious light.— If the tension of the current
used in the preceding experiment be slightly increased without
otherwise changing the conditions, by plunging in saline water the
glass tube traversed by the positive platinum electrode, a brilliant
phenomenon appears at the extremity of this electrode. Not only
does the wire melt, but the glass itself is fused in the heart of the
liquid, giving forth a dazzling light.

136

The extremity of the platinum wire which is in the shape of a
ball becomes centred in a small mass of melted glass and the light
remains very bright during the discharge of the secondary battery,

until the glass, cooled around the electrode, separates it completely
from the liquid (fig. 58).(I)

When using a solution of sea salt in the voltameter, the tension
of more than 300 secondary cells is required to produce this
luminous effect, but if using a solution of nitrate of potash, the
same thing occurs with only 60 or 80 secondary couples.

The manner in which these saline solutions act in presence of
silicate of glass, brought to a very high temperature by the
electric current, is very varied, by reason of the more or less great
fusibility of the silicates formed, as M. Carre has already dis-
covered in mixing different salts with the carbons used for the
ordinary electric light.

(i) Comptes rendus, vol. LXXXIV, p. 914, April soth, 1877.

137

Vitrious light can again be produced by placing the positive or
negative electrode against a glass plate, a small distance above the
saline solution (fig. 59). It is accompanied by a discharge of
white vapour and the glass is at the same time strongly attacked.

This light appears also along the sides of a china basin, as we
have already seen (144) (147) figures 51 and 54. It unites its

Fig* 59-

brilliancy with that of the voltaic arc in M. JablochkofFs electric
candles when the substance which separates the carbons is a piece
of china or kaolin.

The luminous phenomena observed around the glass by means
of induction currents by MM. du Moncel, Gassiot, Grove, &c.,
are also connected with the light in question.

One might be inclined to attribute the brilliancy of this light
to the eombmation of lime,, and silica in the glass; but, if the
spectrum it gives be examined, it is discovered that there are no
perceptible rays, whilst a fragment of calcareous spar placed in the

138

same conditions, though giving forth as bright a light, allows the
characteristic rays of calcium to be seen.

156. The rays of silicium being weak, according to M. Kirch-
hoff's analysis, it is supposed that they do not become visible by
reason of the luminous intensity of the spectrum formed. But
the silicious origin of this light is proved by the important fact that,
in the contact, of the electrode with pure silica, it appears in a

Fig. 60.

state of quartz hyalin crystals (fig. 60). For producing it in this
case, it is only necessary to use, with the same saline solution, a
greater E.M.F. than for glass, about two hundred secondary couples
for instance.

The silica itself being necessarily decomposed by these high
tension currents, the luminous effect results in all probability from
the incandescence of the silicifj"the remarkable similarity of which
to the diamond and plumbago has been shown by M. H. Sainte-
Claire Deville and M. M. Wcehler. To distinguish this light from

139

that produced by an electric current between two carbon points,
we have given it the name of electro-silicious-light.

157. Crowns, arcs, rays and undulating motions.— 1-

If the positive electrode of a secondary battery of four hundred

couples is placed in contact with the moistened sides of a jar of

solution of salt where the negative

electrode is already immersed,

there may be observed, according

to the greater or less distance from

the liquid, either a crown formed

of luminous particles arranged in

a circle round the electrode (fig.

61) or an arc bordered with a

fringe of brilliant rays (fig. 62), or

a sinuous line endowed with a rapid

undulating motion (fig. 63).(l)

Fig. 62. Fig. 63.

(i) Comptes rendus, vol. LXXXII, p. 626, March i3th, 1876.

140

A peculiar noise, continually increasing, is heard, and some
vapour escapes from the water in sharp jets, above the rays of fire,
as if there were a considerable pressure.

If the wire be still farther dipped in, a luminous closed ring
appears ; this ring is succeeded by another, and there is thus a
creation of brilliant waves in the interior of which the liquid moves
with a quick whirling motion.

Sometimes even, round the liquid eddies, little irregular luminous
rings are seen to appear, detached from the glass of the electrode.

If the jar used is a tube like a U, which holds but a small
quantity of liquid, all these waves end by blending in each other,
the liquid becomes entirely luminous and begins to boil violently.

During this time, the deviation of the magnetic needle placed
near the circuit, undergoes continual variations.

158. EleetFO dynamic Whirls. — The following experiment,
which we described in i86o,(l) does not require so great an electric
tension as the preceding ones. It may be made with a secondary
battery of from ten to twenty couples, or with a battery of from
fifteen to twenty Bunsen elements. We will, nevertheless, class it
among those relating to effects of electric currents of high tension,
because it gives results considerably different from those obtained
by employing a much less tension.

The positive electrode is in this case a copper wire, and the liquid
in the voltameter is water acidulated to one tenth of sulphuric
acid. Whilst in ordinary conditions of the electrolysis of water
with this voltameter, by the action of a feeble current, the positive
wire becomes covered with a layer of oxide which slowly dissolves

(i) Bibl. univ. de Gen6ve, vol. VII, p. 332, April 2oth, 1860.

141

in the liquid (8), if a current of a considerable tension be used, a
different phenomenon appears.

The principal oxidation then takes place at the extremity of the

wire. A hissing, similar to that
produced by red hot metal plunged
into cold water, is heard, and the
end of the wire gives forth a jet of
finely divided oxide, which escapes
in abundant flakes and does not
dissolve in the liquid (fig. 64).

At the same time the wire be-
comes sharp pointed, and the inten-
sity of the current which traverses
the voltameter increases consider-
ably.'1'

Fig. 64.

If the pole of a magnet and the

end of the electrode are brought near to each other, the cloud of
oxide moves in an extremely quick gyratory manner, in one
direction or the other, according to which pole of the magnet is
presented. The rotation takes place according to the laws of
Ampere, in the opposite direction to the hands of a watch, before a
north pole (fig. 65), and in the same direction as that of the hands
of a watch, when before a south pole (fig. 66).

The arrows drawn round the whirls indicate the direction of the
gyratory movement under the influence of the magnet, and the
arrows drawn round the magnet indicate the direction of the

(i) The oxide formed in this case appears to be protoxide of copper, rather than
oxide, as we have before said (see note to p. off).

U2

magnetic currents ; B is the north pole, A, the south pole.'

Fig. 45. Fig. 66.

159. This experiment may be arranged as represented in
figure 67. A china or glass basin is placed above an electro
magnet and filled with acidulated water; any kind of wire, in

Fig. 67.
connection with the negative pole of a battery composed of

(i) This experiment is easily reproduced by projection. We have repeated it in that
way at our Lecture at the ''Association Polytechnic" in 1861.

143

fifteen Bunsen elements, is immersed beforehand in the liquid.
The copper positive wire held in the hand, is successively plunged
into the liquid above each pole of the electro magnet.(I)

The cloud of oxide appears, the whirls are developed, and as
the oxide formed does not immediately dissolve in the liquid, but
floats in a very finely divided state upon its surface, the two kinds
of whirls, in different directions, remain some moments traced on
the surface of the liquid after the current has ceased, and even
preserve the motion made by the liquid when under the influence
of the electro magnet

Mr. Sylvanus P. Thompson, has since obtained similar whirls
by making a magnet, excited by an electric current, act upon iron
filings, and has fixed them the same as other magnetic phenomena
produced by electro dynamic actions. (2)

The experiment above described may be compared with several
others on the rotation of liquids traversed by currents in the
neighbourhood of magnets, such as those made by MM. Wartmann,
Jamin, etc. But that which more particularly characterises it is
the curved spiral form of rotation, in consequence of the magnetic
action exercised on the currents radiating round one point formed
by the extremity of the electrode, and the distinctness of these
whirls is all the greater, as the electrode itself supplies, by its
disaggregation, the solid matter necessary to render the motion of!
the currents visible in the heart of the liquid.

160. Crater-like Perforations. — If a sheet of filter paper
moistened with salt water, be put in connection with the negative
pole of a secondary battery of four hundred elements, and if orr

(1) The same current may at once excite the electro magnet and act upon the voltameter.

(2) See La Nature, p. 179, August i7th, 1878.

144

the other hand, the moist surface has just been touched with the
positive pole, there appears below this wire, with a brilliant light
and rise of vapour, a cavity in the form of a crater, its edges
bristling with innumerable filaments, dried up and entangled one
in the other (fig. 68). The positive wire becomes at the same

time covered with a substance
formed by the paste of the
paper deposited upon it; thread-
like residue also adheres to
the electrode over a length of

from 10 to 15 centimetres.
Fig. 68.

The ends of the filaments point towards the positive electrode,
so that, if this electrode be placed below the paper, no crater is

seen to project from the upper
surface, but only a simple excav-
ation, the stringy ledges of which
are as though sucked in, ' and
turned back towards the point
where the positive electricity
passes out (fig. 69).

Fig. 69.

Some filaments, in consequence of their great length and their

-r

Fig. 70.
instantaneous desiccation, become shaped like a hook at the end.

145

Figure 70 represents the details of these electric perforations,
natural size.

These phenomena are another result of the calorific action
exercised by the current, which vaporizes, and instantly dries up
the moist fibres of the organic matter, and are also due to its high
tension, which produces the effects of attraction or suction, and the
mechanical sub-division of the matter subjected to the discharge.

CHAPTER II.

Engraving on glass by electricity,
Other applications.

161. Engraving on glass by electricity.— We have
previously described (152) an experiment in which a glass tube,
with a platinum wire passing through it, serving as electrode to a
powerful voltaic current, becomes instantaneously hollowed in the
shape of a cone or funnel, in the midst of a voltameter containing
a solution of salt. In other experiments (157) on the luminous
effects produced by a current of high tension, against the sides of
a jar of glass or crystal moistened with a solution of sea salt, we
have had occasion to observe that the glass or crystal was strongly
attacked at the points touched by the electrode, and that the
luminous concentric rings formed all around, remained sometimes
engraved on the surface of the glass of the voltameter. We have
further discovered that by using, as saline solution, nitrate of
potash, a much less powerful electric force was necessary than
with chloride of sodium or other salts, in order to produce
luminous effects and devitrification.

These observations have led us to apply the electric current to
engraving on glass or crystal.(I)

162. The surface of a sheet of glass, or plate of crystal, is
covered with a concentrated solution of nitrate of potash, by
simply pouring the liquid on the plate, placed horizontally in a
shallow basin. Then, in the layer of liquid which covers the
glass, and along the edges of the plate, a horizontal platinum wire
is immersed, in connection with the poles of a secondary battery
of from fifty to sixty elements; then, holding in the hand the
other electrode formed of a platinum wire insulated, except at the

Fig. 71.

extremity, the glass covered with the thin layer of saline solution,
is touched at the points where the letters or drawing are required
to be engraved (fig. 71).

A luminous track appears where ever the electrode touches,
and, no matter how quickly they are written or drawn, the
characters become distinctly engraved on the glass. If one
writes or draws slowly, the lines are deeply marked, their breadth
depending on the diameter of the platinum wire serving as
electrode; if it be pointed, these characters may be extremely fine.

(i) Comptes rendus, t. LXXXV, p. 1232, 1877.

us

The wire conducting the current becomes thus transformed
into a graving tool for glass, requiring no effort in the manage-
ment of it on the part of the worker, in spite of the hardness of the
substance which must be cut; for it suffices to touch the surface
of the glass very lightly in order to obtain an ineffaceable
engraving.

The corrosive power is supplied both by the calorific and
chemical action of the electric current in the presence of the
saline solution.(l)

The chemical action of the electric current under these
conditions is very powerful, although exercised upon non-conduct-
ing matter and simply on its surface; it is even more efficacious
in the case of vitreous substances than hydrofluoric acid ; for we
have thus been able to engrave characters on a plate of Sidot
glass(2) which could not be marked by hydrofluoric acid.

Engraving may be done with either electrode; however it
requires a current of less strength to engrave with the negative
electrode, and the engraving is more distinct.

Although these results have been obtained by using secondary
batteries, it is clearly preferable, for continuous work, to use quite
another source of electricity of sufficient quantity and tension;
either a Bunsen battery of a sufficiently large number of elements,
a Gramme machine, or even an alternating current magneto-
electric machine.

(i) The figures produced on glass by static electricity, and the impressions obtained
by M. Grove with inductional electricity, are allied to this corrosion of glass by dynamic
electricity. But as the quantity of electricity supplied by electrical machines or induc-
tion coils, is relatively very small, and as there is besides no electro-chemical effect, such as
that which appears in this case in the presence of a saline solution, these figures and
impressions are scarcely visible. To be perceived they require a. deposit of moisture,
resulting from the breath, which has given them the name of rori^ figures, since studied
by MM. Reiss, Peyre, Wartmann, etc.

(2) This glass is an acid phosphate of lime obtained under special conditions, the discovery
of which we owe to M. Sidot, chemical operator at the Lycee Charlemagne.

149
163. Electric boring- or drilling1.— We think we ought to

point out another application which could be made by the same
means, however difficult its realisation may appear at firs? sight.

We have just seen that, when one of the electrodes which
conducts an electric current of considerable tension is brought in
contact with glass covered with saline solution, it acts as a graving
tool or diamond for tracing lines on the surface of glass
and even hollowing it rather deeply.

Rock crystal may be also attacked, in spite of its hardness, by
the same method; and, if it cannot be engraved so regularly,
at least it breaks into small pieces under the influence of the
electrode and finally becomes disintegrated.

Now, in America, they use at present thin black diamonds for
cutting hard rocks, and for boring wells or mines. (I)

Could not the use of these diamonds, which are very costly (and
gradually waste away by becoming detached from the head
to which they are fastened), be replaced by the action of the
electric current under similar conditions to those which have just
been described, thus obtaining the boring of rocks by electricity ?

Platinum electrodes would not be necessary, because it is not
in this case the metal of -the electrode which is impaired, but the
silicious matter when in the presence of a saline solution. Metallic
points or projections distributed suitably at the end of the boring
bar (a portion of its length insulated and given a rotating move-
ment), would bring the electric current to the surface of the rock
which it would be desired to act upon, and would thus replace
the numerous black diamonds set or inserted in the end of the

(i) See La Nature, August loth, 1878. Le Bondage au Diamant, (L. Bade.)

150

bar, as in the process of diamond boring. The progress recently
made in the production of electricity, by mechanical means,
would facilitate this application.

164. Different applications. — Among the phenomena we
have described in the preceding chapter, there are others, such
as the electro-silicious light, which might also perhaps be made
use of.

If currents of high tension are necessary to discover them,
once known, they become easier to reproduce, either with less
tension, or with greater tension and less quantity of electricity.

It is thus that a considerable number of these phenomena may
be displayed to a certain extent, in a rudimentary state, with strong
inductional coils or even with static electricity. Such is the
phenomenon of the sheaf of aqueous globules, observed with
currents of high tension. A conductor in connection with an
electric machine or with one of the poles of an induction coil,
brought in contact with water, produces a kind of mist, which
might be mistaken for vapour, but that the preceding discovery
proves to be water reduced to an extreme state of sub-division, or
pulverized by the electric discharge. Likewise for crater-like
perforations (160); by studying with a magnifying glass the holes
made by the piercing instruments in static electricity, one finds in
them nearly the same characteristics as in those shown more
clearly by dynamic electricity of high tension.

Applications of the phenomena above described could then be
made with electrical machines or induction coils, when it is not
necessary to have at the same time a great quantity of electricity.

PART.

Analogy of the effects previously des-
cribed with Natural Phenomena. —
Theories which may be deduced
from these phenomena.

CHAPTER I.

Analogy with globular lightning. — On the nature and

formation of globular lightning. — Remarks on

some cases of globular lightning. —

L ightn ing ' ' en chape let.

165. The phenomena which we have discovered with electric
currents of high tension (135-142) present striking analogies with
those of globular lightning and appear to us to be of a
nature to facilitate the explanation of this extraordinary form of
lightning. (I)

(i) " Lightning in the form of a ball, of which we have quoted so many instances, wrote
Arago (Notice sur le Tonnere, p. 219), and which is so remarkable, appears to me at present
to be one of the most inexplicable phenomena in physics."

" P. 396, ibid.— There is but one circumstance in which the physicist does not know how
to engender what nature produces so easily ; he cannot cause lightning in balls ; he does not
know how to produce these spherical agglomerations of matter which move slowly without
losing the property of striking objects. On this subject there is a blank in science which it
would be most important to fill.

152

We have seen in fact that ponderable matter tends to take a
globular form under the influence of a powerful source of dynamic
electricity. We first proved this property on liquids and discovered
liquid luminous globules (135) fig. 39. By increasing the tension,
we obtained, even in the midst of air mixed with vapour from
water, real globules of fire (138) fig. 42 to 47.

We are then naturally led to think that globular lightning must
be produced by a flow of electricity in a dynamic state in which
quantity is added to tension.

Thus, it is in great storms, when electricity abounds in the
atmosphere, and the discharges constitute a kind of powerful electric
current of very high tension, that the lightning appears in a
globular form instead of taking the simple linear shape similar to
that of the sparks of static electrical machines, like that which
happens in storms of less intensity.

166. The nature of globe lightning should apparently be the
same as that of the globular sparks produced in our experiment.

These globes must be formed, in our opinion, of rarified incan-
descent air and gases resulting from the decomposition of vapour
from water also in a state of rarefaction and incandescence.

The water is in fact not only vaporized but decomposed, as has
been shown above (154), at the end of one pole, in consequence
of the extremely high temperature developed by an electric
current of high tension.

167. Although an aqueous surface is not indispensable for
forming luminous electric globules, since we have obtained them
over a metallic surface (146), the presence of water, or of vapour
from water, at least facilitates their formation or tends to give them

153

more volume because of the presence of the gases furnished by
the decomposition of water at a high temperature.

We have more than once observed in our experiments, when
all the discharge is devoted to the production of a single
phenomenon, electric flames in flattened spherical or cup shaped
spherical forms which cover the whole surface of a small jar full
of water(I) upon which a current of high tension has just been
applied.

Also damp air seems more favorable to the production of
globe lightning, and it has often been seen either upon
inundated ground after abundant rain'2' or in an atmosphere
saturated with moisture. Farther on (186-188) we will quote
other examples.

168. We were not considering globular lightning as
enclosing a detonating mixture made of gases formed from the
decomposition of water, and the noise which often accompanies
their appearance as owing to this cause (a noise to which we
shall refer further on) (179). These gases are here so rarified
that they could not produce an explosion ; they are even in an
incandescent state and consequently in a condition altogether
different from that of an explosive mixture, produced when cold,
which would be afterwards suddenly ignited.

(1) This jar was about four centimetres in diameter.

(2) See Arago. Notice sur le Tonnerre, p. 46.

At Massa-Carara, September loth, 1713, during a storm and a deluge of rain, Maffei and
the Marquis of Malaspina suddenly saw on the surface of the paved road a very bright
flame appear of a partly white partly azure light ; this flame seemed strongly agitated but
without any progressive motion ; it dispersed suddenly after having acquired a great volume.

Ibid., p. 50 (Observation made at Trieste in 1841, and sent by M. Butti to Arago) — "the
thunder burst forth at intervals with terrific noise. The street was deserted, for the rain fell
in torrents and the public road was changed into a river; . . .the first thing which struck my

sight was a ball of fire -which proceeded along the middle of the street To give an idea

of the size of this fiery globe and of its colour I can but compare it with the moon ;. .but it
had no precise outline ; it seemed to be wrapped in an atmosphere of light of which it was
impossible to mark the exact limit

154

169. The formation of globe lightning may be explained in
the same way as that of globules of fire obtained in the experi-
ments previously described (137).

The spherical agglomeration of matter subjected to the action
of a powerful electric current is, as we have before said, the result
6f suction or vacuum produced by the passage of the current.

Each of these balls is a kind of electric egg, without the glass
covering, a voltaic spray (139), which the surrounding medium
tends continually to fill; but the abundance of the electric current
rarefies the matter as it flows into the electrified centre. (l)

170. The light of these balls, which is sometimes very dazzling
as many observers have remarked'2' (188), is explained by the
great quantity of electricity in play at the time of their appearance.

The light produced in the electric eggs of the laboratory is feeble
because the quantity of electricity of high tension which passes
through them is very small. But it is known that in the narrowed
parts of the rarefied gas tubes this light is much brighter, and is
more brilliant in proportion as the electric machine or induction
apparatus employed can supply a greater quantity of electricity.

Among the causes of the .brightness of the light sometimes
emitted by globular lightning, may be mentioned the incandescense
of the cosmical particles of the atmosphere which, though in a
very small quantity, add their brilliancy to that of the air and the
gases from rarefied and incandescent water vapour.

(1) The sprays, sometimes spherical; in static electricity and the sparks observed by
M. du Moncel with the induction coil which often terminated in a ball of red fire (Notice
sur 1'appareil d'induction de Ruhmkorff, p. 143) are connected with the same order of
phenomena.

(2) Arago. Notice sur le tonnerre, p. 43 and 46.

155

In fact these cosmical particles contain, besides organic matter,
mineral matter such as iron, silica, lime, &c.,(l) substances gifted
with great radiating power, at a high temperature. Besides the
luminous effects resulting particularly from the incandescence of
silica under the action of electricity have been seen before (153).
(Electro-silicious light.)

171. The colour of the globes, which is very varied, like
that of ordinary flashes of lightning, depends, in our opinion,
on the hygrometrical condition of the atmosphere and also on the
quantity of electricity in play.

If the water vapour is very abundant, the hydrogen proceeding
from its decomposition predominates and the ball inclines to a
red colour, as that is the colour most peculiar to rarefied hydrogen,
when traversed by a strong current.

If, on the other hand, the electric current is relatively less
abundant, rarefaction and decomposition are less complete in its
path and the colour inclines towards the violet blue peculiar to
rarefied air.

The intervening shades would be explained by the varying
proportions between the rarefied gases of the air and of water
vapour.

172. The peculiar odour which accompanies the fall of globes
of fire, and even of ordinary lightning, may again be explained
by the combustion of the cosmical particles joined to that of the
matter itself directly struck by the discharge at the point where it
reaches the ground. It can be understood that, in the long path
of a flash of lightning or in the passage of a column of air by

(i) See Les poussiere de 1'air, by Gaston Tissandier, p. 12, Paris, 1877

156

the electric current, the cosmical particles encountered would be
of sufficiently great number to give forth a perceptible odour by
their combustion.

The ozone and nitrous products formed by the combination of
the elements of air also doubtless contribute a considerable
portion.

173. The noise which accompanies the appearance of globe
lightning and which is noticed in the experiment described
above (135), proceeds from the rapid vaporization that the
electric current developes.

We will add that, during the discharge produced in particular
by the positive electrode over the distilled water (140), a very
marked sound of blowing is heard, evidently owing to the
vaporization of the water which heats under the action of the
flame emanating from that electrode much more strongly than
under the action of the negative electrode.

174. These considerations, joined to that of the ordinary
direction of positive atmospheric electricity, incline us to think
that the electric sign of balls which result from the direct
flowing of electricity from the clouds should be positive, whilst
that of the fires of St. Elma, sprays, luminous columns (189),
and other electrical effects produced by induction, must be
negative.

175. The gyratory motion sometimes observed in globe
lightning would simply result from the reaction due to the flowing
of the electric current; the same as in the case of the spiral
motion of liquid globules (136), or luminous fibres, composing the
ovoid flames (138) formed on the surface of the voltameter.

157

176. Globular lightening appears either in the shape of a
simple fall of balls of fire, more or less numerous, which disappear
immediately, or in the shape of a single globe which moves slowly,
and remains visible for some time.

In the first case, the globes of fire appear to owe their origin to
flashes of lightning of a peculiar kind which we will describe
further on by the name of "lightning en chapelet" (188), and
the formation of which we will explain.

They are flashes of lightning produced by the flowing of a
greater quantity of electricity than that of ordinary lightning and
involves the production of spherical agglomerations of rarefied
electrified matter in their path.

177. The second case, consisting of the slow passage of a
fulminating ball, may be produced in our opinion in two different
ways.

We have before pointed out (141), that the globules of fire
obtained over water, or even over any kind of conducting surface,
by means of an electric current of high tension, naturally followed
the movements of the electrode at the extremity of which they
appear ; so that, if one works in the dark or if the wire serving as
electrode and oscillating like a pendulum be concealed by a
screen, only a globule of fire is seen moving above the
conducting surface.

So, in nature, if a storm cloud charged with a great quantity of
electricity happens to pass at a short distance from the ground it
may form a column or water spout of moist air^ invisible^ and
strongly electrified, which serves as electrode^ and produces the flow
of the electric current in the shape of a globe of fire which

158

appears at its extremity. This column being essentially mobile,
the globe of fire naturally follows all its movements.

178. But the slow motion of these globes may be also
produced in another way, although there may be no displacement
of a column of moist electrified air.

We have shown in the experiment of the wandering electric spark
(142) fig. 48, how, in certain conditions, a globular spark can
move spontaneously and slowly enough for the successive devel-
opment of its capricious sinuosities to be seen. It was only
necessary to cause the production of a spark of dynamic electricity
of high tension between the two armatures of a condenser, the
very thin insulating plate of which could easily be pierced where
any crack previously appeared. The spark, instead of being of
instantaneous duration, wanders along, burning before it the sub-
stance even of the condenser and collecting it in a globule of fire.

It may then be admitted that there are formed in the atmosphere,
at the point where globular lightning appears, the elements of a
condenser in which a layer or column of damp air strongly
electrified plays the part of upper armature, the ground the part
of lower armature, and the layer of air interposed, that of the
insulating plate. (I)

This layer of insulating air being passed by the electric current,
the flow appears in a globular form between the ground and the
column or damp electrified layer forming the upper armature.

(i) The intervention of a strip of insulating air, in the production of the phenomenon
has been, besides, already admitted by M. du Moncel : "The slow movement of the
ball of fire," says M. du Moncel, "would only be the result of the variations in the
direction of this insulating band, or in the current of air which would have occasioned it,
variations which would displace the point where the flow of the electric fluid would appear
in a luminous state." (Notice sur le tonnerre et les eclairs, p. 51, Paris, 1857).

159

When the base of this column presents a considerable area, as
sometimes happens if it forms a part of the electrified cloud
approached very near to the ground, the globe of fire remains in
connection with this armature without changing its place, and
continues its way alone, passing through the layer of insulating air
in an irregular manner, according to the variations in thickness or
the resistance presented, just as, in our experiment, the little ball of
fire makes its way between the upper and lower armature of the
condenser without displacing the electrodes or the armatures.

In the experiment we recall, the spark is, it is true, a globule
of solid matter in fusion ; but the other examples we have quoted
permit us to suppose that spheroids of incandescent gas would
present the same phenomena.

The objection may be raised to these comparisons that the
natural balls of fire do not appear at the extremity of metal
electrodes. But the identity of lightning with sparks from
electric machines is now admitted ; and although these sparks come
from metallic conductors the masses of water vapour which form
electrified clouds are considered as similar conductors. A
metallic electrode conducting a current of high tension may then
be compared with a column of damp air by which the electricity of
storm clouds sometimes descends nearly to the surface of the
ground.

179. By the preceding considerations it may be explained how
the globes sometimes disappear without noise and, in other
cases, are accompanied with thunder.

When the thickness of the layer of insulating air which
separates the electrified layer of cloud from the surface of the

160

ground becomes too great, in the path of the fulminating globe,
and when, on the other hand, the quantity of electricity supplied
by the storm cloud does not increase, the electric flow ceases and
the globular flame disappears silently, just as the globule of fire
produced on the surface of the condenser ceases to appear when
the thickness of the insulating plates becomes too great.

180. If, on the contrary, the storm increases in intensity or
(the electrified cloud approaching nearer the ground) fresh
quantities of electricity arrive on the surface of the layer of
insulating air, the flow, instead of continuing to progress in a
relatively calm and silent manner in the globular form, strikes
sharply like the ordinary discharge accompanied with the noise
of thunder.

It can then be understood that, from the point itself where the
lightning globe appeared, zigzag or sinuous darts of lightning,
which strike the surrounding objects,'1' flash in all directions.

181. But we do not at all mean by that that the noise is due
to the explosion of the fulminating ball itself.

If reference be made to the ideas we have given upon their
nature, according to the analogies drawn from our experiments,
it will be understood that a small mass of air, rarefied and
luminous from the passage of the electric current, cannot burst
with the noise of thunder and disperse in fiery darts.

The source of the final phenomenon is in the reservoir itself
of electricity that the storm cloud encloses and which is

(i) Our explanation on the point is nearly identical with that given by M. du Moncel :
"The explosion of the ball of fire and the flashes of lightning which it sends forth laterally"
said M. du Moncel "would be nothing more than the electric discharge pure and simple,
determined by the conducting substances interposed in the band of insulating air in the
range of which the meteor would happen to be." (Notice sur le tonnerre et les eclairs, p. 51.)

161

discharged at the instant when the flow had begun in the form
of a globe of fire.(I)

182. When globular lightning appears in the form of a fall of
fiery balls which are seen only for an instant (176), the thunder
which accompanies this fall should not be attributed to the balls
themselves but to the whole of the lightning en chapelet from
which they are derived (188), and of which they constitute
detached grains.

183. The exceptional intensity of the thunder, often mentioned
in connection with the fall of globular lightning, is again
explained by the quantity of electricity in play, always greater in
the manifestation of this phenomenon than in ordinary cases.

The volume of electric fluid, if it may so be called, that is to
say, the mass of ponderable matter passed through and rarefied
by the discharge, is then greater. Hence, naturally, a greater
vacuum is produced.

But how can electricity produce a vacuum ? has been asked all
this time. Our experiments permit us, we think, to simply reply :
by powerful and instantaneous calorific action which developes
electricity and vaporizes all matter placed in its path.

The greater part of the phenomena that we have described
(135 to 1 60), is, in fact, only the consequence of the vaporization
produced on liquids or humid surfaces by an electric flow in
which quantity and tension are both united.

(i) We thought at first, like several authors or observers, that the explosion accompanying
the disappearance of the fulminating balls, was due to the electricity accumulated in them,
but we have modified our ideas on this point since our last experiments. There is doubtless
accumulation or agglomeration of electric matter, since there is spherical enlargement of the
space rendered luminous by the flow of electricity, but, as we have before stated, it is the
electricity pouring from the whole cloud which produces the discharge, accompanied with
the noise of thunder, and not the small quantity of electricity in the ball itself,

162

184. The reason lightning conductors have often proved
ineffectual in cases of globular lightning is explained by
considering that the appearance of a fulminating ball reveals the
commencement of an abundant and continuous flow of electricity
from the storm cloud in a particular spot chosen, and that the
simple action of influence exercised by the proximity of a lightning
conductor would not be able to stop this flow when once begun.

If these slowly moving balls appear harmless in themselves,
since observers near to whom they have sometimes passed have
received no hurt, they are not the less a source of great danger;
for they represent either the extremity of a cloud electrode or the
chosen point where it exercises its greatest influence and they
forebode an imminent discharge, all the more destructive because
of the greater quantity of electricity in play.

The directions in which a lightning conductor can be effective,
cannot be too numerous to prevent the formation of this particular
point from which an abundant electric discharge may take place.

MM. Melsen's and Perrot's many pointed lightning conductors
appear to us specially designed for this object and preferable to the
single pointed ones of great height.

A few of the experiments that we have described, such as the
spark in form of a flame pointed basket, produced below a positive
electrode conducting a current of high tension to the surface of a
liquid (143), fig. 50, the perforations produced on damp organic mat-
ter with crater-like formation, bordered with filaments turned back
and diverging round the point struck by the discharge (160), fig. 68,
indicate the form that negative electricity seems to prefer when
going to meet, as it were, the positive electricity and neutralizing it.

163

These experiments are then in favour of the arrangement of
a basket of points, adopted by M. Melsen's for the lightning
conductors of the 1' Hotel de Ville at Brussels, and strongly
support the views of the learned Belgian on this subject.(I)

185. The experiment that we have just referred to (160),
fig. 68, besides offering a striking picture of the effects of
desiccation produced by lightning upon vegetables and of their
division into laths, thongs, or innumerable shoots, explains how
they are torn to pieces, or uprooted, and the effects of suction
which often accompany discharges of atmospheric electricity.

186. Instance of globular lightning- at Paris, in 1876.

The preceding explanations appear to agree satisfactorily with
the facts that we have had opportunity of gathering or of personally
observing concerning globular lightning.(2)

The conditions before pointed out (165-167) as being favour-
able to the manifestation of this phenomenon ; that is, the presence
of a great quantity of electricity in the atmosphere, constituting
by its frequent and continuous discharges a kind of dynamic flow,
joined to the production of abundant rain saturating the air with
water vapour, were realized on the occasion of two violent storms
which visited Paris on July 24th and August i8th, 1876.

It was also proved that lightning fell in several places in a
globular form.

On the 24th of July, between half past three and four in the

(1) Des Paratonnerres a pointes, a conducteurs et a raccordements terrestres multiples,
par Melsens'. Bruxelles, Hayez, 1877.

(2) Comptes rendus, vol. LXXXIII, pp. 321 and 484, July 3ist, and August 2ist,
1876. — La Nature, 4th and 5th years. September 3oth and October 28th, 1876,
April 7th, 1877.

164

afternoon, a deluge of rain mingled with large hail and accomp-
anied with lightning and thunder poured down upon the Place
de la Bastille which we were at that moment crossing. There
being comparitively little wind, the storm cloud remained nearly
stationary for some minutes ; the discharges were incessant, and
several claps of thunder, following the lightning without any
appreciable interval, announced that the latter had struck the
earth several times in the neighbourhood.

Making immediate enquiry, we learned that the lightning had
just fallen three times running nearly at the same point on the
theatre Beaumarchais, in the court and garden of the house
No. 28 in the rue des Tournelles, known at the Marais by the
name of the hotel de Ninon de Lenclos.

The manager of the theatre, who happened to be in the dress
store, a little room situated in the upper part of the building,
had seen a shell of fire fall about the size of a hand.

In the rue des Tournelles, a workman living on the fourth
storey had seen a globe of fire, about the size of a cannon ball
pass along the edge of the roof near to a pot of flowers, only
breaking one stem and then falling in the court yard. At the same
moment, another workman on the ground floor saw three little
balls of fire above the ground of the same court yard, which was
at that moment completely flooded.

Over the way, M. Languereau, manufacturer of bronzes, saw in
his garden two or three incandescent particles fall, without any
clearly defined outline, and drown themselves, to use his express-
ion, in the garden which was transformed into a vast basin

165
by the amount of rain, falling like a veritable water spout.(I)

187. These incandescent particles would not be formed, in our
opinion, of ignited matter, but principally of rarefied air and gases
from luminous water vapour, Hke the globes and electric flames
produced in our experiments.'2'

However, as hailstones are found having inside them kernels of
mineral or organic matter, there may be also found in this kind of
luminous particle some cosmic corpuscules borrowed from the
atmosphere.

188. Lightning "en chapelet"— The storm of August

1 8th, 1876, was even more remarkable than the preceding one for
the intensity of its electric phenomena. It happened after a long
period of heat and dry weather and was accompanied by torrents
of rain. This storm, the different phases of which we attentively
followed from one of the highest points in the neighbourhood of
Paris (the heights of Meudon) where we happened to be at that
season of the year, gave us the opportunity of observing a
very rare kind of lightning, not classed in meteorology,
which appeared to us of a nature to throw new light on the
formation of globular lightning.

The storm broke towards six in the morning in the neighbour-
hood of Paris. A vast cloud darkened the sky and gave forth a
series of flashes of lightning of great length and very varied shape:
some were forked, others presented curves with numerous points

(1) The material damages were insignificant, as might be expected from the fall of this
column of water, which would easily conduct the greater part of the electric current to the
ground. A piece of xinc from the roof of the theatre, lifted up and shot upon the next
house, the ignited gas at the end of a lead pipe, and some shocks felt by different persons
who witnessed the phenomenon ; such were the only accidents heard of.

(2) We have had, in fact, often occasion to observe, that the least cause, such as a breath
or draught of air, was sufficient to change the spherical shape, and alter the outline of these
flames.

166

or with unbroken outline. One of them, doubled upon itself, was
exactly similar to a curve known by the name of the "folium de
Descartes."

This lightning appeared to be principally composed of brilliant
points similar to the tracks of fire produced on a damp surface by
an electric current of high tension.

Towards seven in the morning, at the moment when the storm
began to extend over Paris, a very remarkable flash of lightning
shot from the cloud to the ground, describing a curve similar
to an elongated S, and remained, visible for an appreciable instant,
forming a chain of brilliant beads scattered all along a very narrow
luminous thread (fig. 72).

Fig. 72

This flash appeared to strike Paris in the direction of Vaugirard.
The newspapers, in fact, said a thunderbolt had fallen at Vaugirard,

167 %

Crenelle, cS^c., and further, that it had been seen in an ovoid or
globular form.'11

It is probable that the fall of the lightning must have happened
simultaneously at different points, and that it divided into several
branches or beads when near the ground ; for we had only seen
one flash reach the earth in this direction.

The rain was very heavy, so that the air through which the
discharge passed must have been entirely saturated with water
vapour.

The lightning also fell during this storm in a globular form(2) on
a house 35, Rue de Lyon. This fact was also r.^ntioned in
the newspapers, (3) and we ascertained, by enquiry, that it was
true.

(1) We extract from newspapers published in Paris, Saturday, August igth, 1876, the
following passages :

" The long expected storm has at last arrived. Towards midnight flashes of lightning
began to line the cloud, silently, yet increasing every instant in intensity. Towards four in
morning, they followed each other uninterruptedly, like sparks from fireworks.

It was remarked that the thunder claps differed from the ordinary roar. They were not
all like the usual crackling, but a series of dull sounds, similar to a cannonade. . .

Lightning has fallen in several places and produced rather curious phenomena.

Thus, on the boulevard de Vaugirard 259, the electric fluid entered by the chimney, cross-
ed a room occupied by a servant, who was fortunately absent, and after having set fire to a
bag of linen, left the room, breaking two panes of the window in its way.

. . . Nearly at the saute instant a flash struck the house, No. 99, rue d'Assas. The
fluid, in an OToid fortn, destroyed the west gable of the house and shot it to a great dis-
tance in the surrounding gardens. Pieces of stone from the cornice, in falling on the
balcony of the fifth storey, sent forth thousands of sparks. The zinc which bordered the
gable was cut as though with shears.

(2) It is probable, says M. H. de Parville, that the Phenomenon of globe lightning
appears more often than one thinks ; it may have, until now, escaped the notice of observers
not prepared for it. Thus, according to M. Alluard, director of the Observatory of Puy-de-
Dome it is not uncommon to see, during a storm, quantities of little balls of fire rebound on
the side of the heights.

(3) We also extract this passage : "The storm which yesterday, August i8th, broke
over Paris, was accompanied at moments by torrents of rain. At No. 35, Rue du Lyon ;
lightning appeared in the form of a luminous globe. It also fell in the court-yard of some
works, 7, Rue Jules-Cesar (a few steps from 56, Rue de Lyon).

168

Among other witnesses, a pupil in the chemist's shop situated
on the ground floor of this house declares that he saw fall at
some yards distant, at the same instant, two balls of fire so
bright as to dazzle him, which disappeared on reaching the
ground.

Although, from Meudon, we could not see the lightning which
struck this point in Paris, on account of the curtain of rain which
hid it from view, observation of the lightning " en chapelet "
which fell at Vaugirard permits us to think that that of the Rue
de Lyon would be of the same kind. In short, those which
appeared in the midst of the clouds, as we have already mentioned,
presented the appearance of a series of brilliant points rather than
of luminous unbroken lines.

189. The quantity of electricity spread through the atmosphere
was so great during this storm, that very curious effects of influence,
similar to the sprays and St. Elmo fires appearing on the ends
of masts of ships, were observed by M. Trecul in the neighbour-
hood visited by the thunder'1'

190. This kind of lightning appears to us to constitute an
indicative phenomenon'2' showing the transition from the ordinary
form of lightning, in sinuous or rectilineal lines, to the globular

(1) "During the storm which happened on the morning of August i8th," said M. Trecul
" between seven and eight o'clock I was writing before my open window. Tremendous
claps of thunder which seemed in the immediate neighbourhood were heard repeatedly.
At the moment the nearest were heard, or nearly at the same time, small luminous columns
fell obliquely on my paper. One of them was about two metres in length— in appearance
they resembled lighted gas with an indistinct outline ; no detonation took place, only when
nearly extinguished, they left the paper with a slight noise." (Comptes rendus, vol.
LXXXIII, p. 478, August zist, 1876).

(2) The form is much more marked and clearly shown in .some cases than in others ;
these special cases are those in which the nature or the form is less hindered or constrained
by other influences or confounded with them. We call these cases "striking and indicative
cases." Bacon, novum organum, lib. II, §20.

169

form. It can be understood that the beads of lightning might
acquire a certain volume and become globes of fire.

We concluded from this observation that globes which fall
in a greater or less number, accompanied by thunder, and
which disappear immediately, could be considered as being
derived from lightning "en chapelet."

191. This formation of luminous grains, alternating with darts
of fire, would be the result of the passage of electricity
across a ponderable medium, and may be compared either with a
string of incandescent globules (presented by a long wire
melted by a voltaic current, the extremities of which remain for
an instant suspended in fusion at the poles of the battery), (99)
fig. 23, or to the expansion resulting from the flowing of the
whole liquid vein.

Electrified and luminous matter must naturally be dissipated
more slowly, in such agglomerations, than the line which unites
them, and the longer duration of the lightning may be thus
explained.

192. This observation agrees with another of the same kind,
quoted by M. du Moncel in the description of a series of flashes
of lightning with tracks of long duration. During a storm in
London on the night of the igth of June, 1857, several flashes
were remarked " which lasted some moments and did not
disappear before being, as it were, melted into beads of light. "(l)

193. We have thus been led to propose collecting these
examples of lightning of a peculiar character and to classing them
under the name of lightning "en chapelet" among the meteoro-
logical phenomena.'2'

(1) Notice sur le tonnerre et les eclairs par le comte du Moncel, p. 54.

(2) Comptes rendus, vol. LXXXIII, p. 484, August 2ist, 1876.

170

194. Since then, several witnesses have confirmed the fact of
the existence of this kind of lightning.

In a communication addressed to the Academic des Sciences,
November 2oth, 1876, M. E. Renou wrote saying that our
observation reminded him of a case quite similar which he had
witnessed before.

"During a violent storm which broke out on the evening of July soth, 1859, a* ponts de Brave,
commune de Souge, on the boundary of the departments of Sarthe and Loir-et-Cher, the
lightning, said M. E. Renou, seemed to fall on the Italian poplars situated on the banks of
the Braye, between 200 and 250 metres from where I was; the lightning made a vertical track,
slightly sinuous, formed of balls nearly in contact, exactly like a string of beads and extremely
brilliant."(i)

M. Renou also brought forward another argument in support
of the explanation we have given of the origin of globe
lightning, by comparing the diameter which the grains of bead
lightning appeared to him to have at a certain distance, with that
of the fulminating globes seen close to by some persons.

" This appearance," said M. Renou "was instantaneous, but, according to the impression
which it left me, I estimated the diameter of the balls at the tenth part of the diameter of
the sun ; an angle of three minutes to 200 or 300 metres would give to these globes a diameter
of om.2o ; it is the same diameter which has been given to those globes of fire occasionally
seen to pass slowly through the interior of houses without hurting any one present."

195. The R. P. Van Tricht(2) relates how, during a violent
storm accompanied by hail, which took place at Namur, July 24th,
1877, and which was considered the worst in that month, one of
his colleagues, watching with him, "very distinctly perceived one
of those flashes of bead lightning spoken of in the Comptes
rendus de 1' Academic des sciences at Paris.

196. On the other hand M. Daguin, Professor at the Faculte
des sciences de Toulouse, wrote to us lately :

(1) Comptes rendus, vol. LXXXIII, p. 1002, November 2oth, 1876.

(2) See 1'Etude des orage en Belgique, par A. Lancaster. Annuaire de 1'Observatoire
Royal de Bruxelles, p. 279, 1878.

171

" In confirmation of the form of lightning in beads, of which you have quoted instances, I
can tell you of a flash of that nature which I saw pass from the clouds to the earth. I was
then at the Toulouse Observatory and I had classed this case among the strange and varied
forms often presented by lightning seen from an elevated position."

197. Other instances of similar flashes of lightning have been
more recently published in England :

" On the evening of August i6th, 1877," wrote M. B. Joule (i) "a violent storm took place
at Southport. — Among the most brilliant flashes which I observed was one which presented
a phenomenon I had never before witnessed. From the point where it left the clouds to its
fall into the sea it seemed formed of little detached fragments which gave it the appearance
represented below (fig. 73).

73-

198. This latter observation has been supported by another
published in the same collection. (2)

Mr. E. J. Lawrence writes— "I can confirm this fact, that lightning sometimes presents a
punctuated form. . . . About forty years ago, during a storm, accompanied by heavy rain(
which I witnessed at Ampton (Suffolk), flashes of lightning followed each other incessantly

(1) On a remarkable flash of lightning ; Note read by M. B. St. J. B. Joule at the
Society of Physics, &c., Manchester. " Nature," vol. XIII, p. 260, July 4th, 1878.

(2) "Nature" vol. XVIII, p. 278, July nth, 1878. Remarkable form of lightning by
E. J. Lawrence.

172

for more than half an hour, and about a quarter of them, as far as I can remember, presented
this peculiar appearance. Since that time, I have often watched for it, but I have only seen
it once more, and then there was but one flash of this kind, among a great number on both
occasions, these punctuated flashes were of excessive brilliancy, and presented a curved
sinuous form, without sharp angles ; one among them had the shape of a nearly perfect fig. 8.

199. The reality of the existence of bead lightning, or simply
punctuated (when seen at a greater distance), appears to us
demonstrated by the preceding facts, and allows us to form them
into a separate class to which we may call the attention of observers.

It would be further interesting, in the case of observing more
lightning of this nature, to ascertain if it has been followed by
the fall of globular lightning, which might confirm the views we
have just expressed.

CHAPTER II.

Comparisons with the phenomenon of
Hail. — Upon the formation of Hail.

200. Mechanical and calorific effects upon watery substances,
or moist surfaces, by discharges of dynamic electricity of high
tension (143-145), may possibly throw a new light upon the origin
of hail.(I)

The phenomenon of the spray of aqueous globules (figs. 49
and 50) (143), which is projected or thrown forth when a powerful
electric current happens to strike the surface of the liquid, shews
that a similar effect may be produced when an electrified cloud,
or an atmospherical electric current, penetrates another cloudy
substance in either its natural state, or less powerfully electrified.

It is true that clouds are not exactly masses of liquid, but those
in very elevated regions are composed, we know, of very fine and
light ice crystals, the cohesion of which is less than that of
ordinary ice, and which may be considered as almost equivalent
to a liquid mass in suspension in the atmosphere. It may then

(i) Comptes rendus, t. LXXXI, p. 616, et t. LXXXII, p. 314, October n, 1875, and
January 3ist, 1876.

174

be conceived that electrical discharges may, in such a case, result
in a similar effect to that which they produce on a liquid, and that
the water of these ice crystals, melted and split up at the points
where the discharges take place, may be ejected in the form of a
spray of globules, as in the experiment.

Moreover, on account of the low temperature of the bulk of
the cloud itself, or of the great altitude where the phenomenon
takes place, these globules may be instantaneously frozen, and so
originate hail stones.'1'

201. According to the greater or less density of these clouds,
and the quantity of electricity in play, the calorific or mechanical
effects produced by the electric current may alternate, become
mixed, or exchanged the ones for the others, the same as was seen
in the experiments already described, when the electric current
caused either a mechanical effect, such as the expulsion of water
in a liquid state, or calorific effect, shewn by an abundant
emanation of vapour, according as to whether it met with a
quantity of water, or only a moist surface (144) fig. 51.

When the calorific effects dominate in the action of an electric
current in the centre of a cloudy mass, the water is no longer
only scattered, but vaporized by the current, and this vapour,

(i) If the experiment was effected with a much higher tension upon ordinary water,
under a receiver, at a very low temperature, the drops emitted would be evidently solidified,
and a more complete artifical re-production would be made of the natural phenomenon.

The difficulties in carrying out this experiment being considerable, on account of the space
necessary for the enclosed cooled atmosphere, we made an analogous experiment by operat-
ing in the ordinary temperature, and taking a concentrated solution of salt (nitrate of
potassium), heated nearly to boiling point, so that the drops thrown forth by the electric
discharge might be quickly solidified by refrigeration.

The electric current, being brought to the surface of this solution, contained in a cup
placed on a pedestal, about two metres from the ground, so as to allow a considerable height
for the drops to solidify in falling, the spray is produced, and we thus obtain by electrical
means an artificial hail of nitrate of potassium.

175

immediately condensed in liquid drops at the heart of the cold
cloud, might still give rise in this case to hail stones.'1'

202. We are then led to consider hail as arising from the
freezing of the finely divided cloud water, vaporized by electric
discharges in the lofty and cold regions of the atmosphere.

203. The intensity of the electrical phenomena generally
shewn in hail storms, when the flashes of lightning are almost
incessant and resemble the continuous discharge of a powerful
current of dynamic electricity at a very high tension, shews the
importance of the part that must be played by the mechanical
and calorific effects which come into action in the production
of hail.

On the occasion of violent hailstorms which took place in
Switzerland and France, on the yth and 8th, July, 1875, there
were from eight to ten thousand flashes of lightning per hour,
resembling an immense conflagration.(2)

An idea may be given of the enormous quantity of heat and
water vapour that can be produced by such a torrent of electricity

(1) It seems at first sight that a substance so finely divided as water vapour could only
produce extremely small hailstones. But it must be taken into consideration that, before
congealing, the vapour necessarily passes through condensation, and that if it is produced
rapidly and in great abundance, it may be quickly condensed in drops of considerable size
against the cold parts of the cloud, just as it is condensed in drops against the interior
surface of the lid of a vessel full of water at boiling point.

(2) Upon two hailstorms, &c., by M. D. Colladon (Comptes rendus, t. LXXXI, pp. 104,
446 et 480). "The electric phenomena," writes M. Colladon, "were very remarkable
about the central portion of the hail cloud ; the flashes of lightning followed each other
with such rapidity, between midnight and one o'clock, that we counted on the average two
to three flashes per second, which would make eight to ten thousand per hour,— at every
place visited by this storm the lightning has been compared with a gigantic conflagration,
so permanent did the glare appear."

During a violent hailstorm which took place on the 25th July, 1877, in the High Alps, —
" the lightning lit up the sky with an uninterrupted light ; the thunder claps followed each
other without interval." Note de M. Le Capian. Bulletin de 1' Association scientifique de
France, p. 367, September gth, 1877.

176

in the heart of the clouds when we see the amount of vapour
liberated in the experiments given above.

204. The violent oscillations produced in the middle of clouds
from which hail is falling, and the rapid transformation of the
cirrhus into nimbus also support this view ; for the suddenly
appearing nimbus can only arise from rapid vaporization, and
from water condensed from a portion of the cirrhus.

The multiplication of sifts in hail clouds, and their distorted
form, are likewise explained by the effect of electrical discharges,
if we compare them with the effects produced by high tension
currents upon moist substances (160).

205. The high wind, which nearly always accompanies hail
storms, may be attributed to the rarefied air caused by the electric
current quickly vaporizing the moisture which it meets with in its
passage, and to the inflowing of the surrounding air, which
immediately fills up the vacuum formed.

206. The noise which precedes or accompanies the fall of hail
is due to the electric current penetrating the cloud, and
consequent disintegration or vaporization resulting, just as the
noise produced by the passage of a high tension current in a liquid
or upon a moist surface, is due to the projection in the form of a
spray of the water globules, or to the rapid issue of jets of
steam (144 to 157).

207. The lightning which accompanies hail storms, with or
without thunder, arises from the fact that, in this collision between
two moist masses possessing great mobility of form, sometimes
the one penetrates more or less deeply into the other, just as, in
the action of a high tension current upon the surface of a liquid,

177 ,

the emanation is produced in the form of luminons rays, accom-
panied with a slight noise, when the liquid is rendered strongly
negative by the electrode being completely submerged, whilst
loud sparks appear if the liquid be, on the contrary, strongly
positive, and the discharge takes place at the negative electrode
(135 to 136).

208. In the same way it may also be explained how hail can
be produced without apparent electrical manifestations and yet
owe its origin to the presence of electricity.

We have, in fact, obtained in our experiments the production of
vapour, even without luminous phenomena, when the quantity of
electricity given by the high tension current is very small.

In the same way, without any visible lightning or perceptible
noise of thunder, there may be a production of vapour in the
atmosphere, and solidification in the cold regions, in the form of
small hail when there is only a small quantity of electricity.

209. The length of time occupied by the fall of hail at any
one point being sometimes very short, may be explained by the
brief duration of the electric discharges, (I) and by the strong wind
which accompanies the storm cloud, and rapidly drives it towards
other points.

210. The fall of hail in paths sometimes so narrow that in the
same place and in the same quarter of a town points only a very
short distance apart shew no sign of it when the portion between
is alone affected, (2) may be explained by the vaporization and

(i) We have been able to point out at Paris, where hail storms have generally but a very
short duration and a relatively low intensity, falls of hail which only lasted from half a
minute to a minute, and followed almost exactly the appearance of lightning and claps of
thunder (28th March, and 3rd and 25th May, 1876).

(2) This was also pointed out at Paris in the course of the year 1876.

178

solidification of the water around the tracks made by the lightning,
always much longer than they are wide.

As to the long and wide belts of hail which cover a great extent
of country, these may naturally be caused by the passage of the
cloud itself under the action of the wind accompanying it.

The width of the belt corresponds with that of the cloud or
group of clouds, and its length with the distance traversed.

211. The tracks of rain comprised between two belts of hail
may result from the interior mass of the cold clouds, through
which the discharges take place, being re-warmed by the frequency
of the flashes of lightning, the scattered or vaporized water is
only condensed in the form of rain in the intermediate part, whilst
solidification may still occur on the borders and continue through-
out the whole of the passage.

212. The intermittant periods and returning force of the
storm which may be remarked either in the fall of the hail or in
the gusts of wind which accompany it, are quite analagous to those
noticed in our experiments when the electric current debouches
upon a moist surface (145), and may be explained in the same
manner.

When the electrified cloud has reduced in the form of vapour a
portion of the cirrhus into which it penetrates, there is a
moment before it can meet with a new substance to vaporize ; but
the remainder of the cirrhus immediately fills the vacuum
formed ; a fresh discharge is produced, and, consequently, another
emission of vapour, or of finely divided water, and formation
of more hail stones.

179

213. The ovide or pointed form of the hail stones, and their
sharp corners or protuberances, may be attributed to their electric
origin; for in the experiment of the "spray" (143 to 144), the
globules have also an ovide form, and the spark whence they are
projected has the appearance of a crown of naming points.

In other experiments, where the current acts upon a moist
paste, which preserves the shape arising from the action under-
gone, craters may be observed bordered with many pointed
protuberances exactly characteristic of the passage of the electric
current (160).

214. The light sometimes emitted by hail stones is very likely
due to electricity. Although, in our experiments, we could not
distinguish whether the aqueous globules have their own light or
partake of the reflection of the spark, it is probable that the
electric flow imparts to them a brief phosphorescence, since, with
a greater E.M.F., the moist air itself becomes incandescent (138).

215 The interior structure of hail stones varies, as we know;
some present a radiating structure and seem to have been formed
from a single jet, others have an opaque white centre, covered with
layers of ice, alternately opaque and transparent.

The formation of the former may be explained, as we have
already described (200), by the production of an electric spray of
aqueous globules, immediately solidified in the volume they possess
at the moment in which they are emitted, or by the water vapour
produced under the calorific action of the discharges, condensed
in large drops and immediately frozen within the circle of the
cloud at a low temperature (201).

In our experiments, the greater the quantity of electricity given
by the high tension current, the greater are the globules emitted

180

so, in nature, the largest hail stones are produced in the storms
where the electrical manifestations shew the greatest intensity.'11

216. The structure of the hail stones formed of alternately
opaque and transparent layers seems to prove a progressive
development within the body of the clouds.

This increase of volume has been attributed to various causes,
all of which may be taken into consideration: — either to an
oscillating motion in the hailstones, as supposed by Volta, when
the arrangement of clouds was favourable ; — -or to their fall
across a great thickness of clouds, as presumed by M. 1'Abbc
Raillard'2' ; — or, according to MM. Saigey, Daguin, Fron,. Faye,
Secchi, &c., to their prolonged spiral motion under the influence
of the whirlwinds which usually accompany hail storms, and which
have been observed by Lecoq and by MM. de Tastes, Severtzow,
Buchwalder, &c.

These whirlwinds may be able to suspend the hail stones for a
certain time in the clouds, and so help to increase their volume.

But, in order to explain the alternately opaque and transparent
layers, we thought that might be accounted for by successive
vaporizations and solidifications, added to the spiral motion. (3)

The opaqueness of the cloudy centre which forms these hail
stones, seems, in fact, to prove the sudden chill and solidification

(1) In the storms in Switzerland, already quoted, " the size of the hailstones," writes
M. E. Plantamour, " attained proportions unusual in our latitudes, and a strong west wind
transformed them into veritable projectiles, breaking everything in their passage ; some of
the size of a walnut, a pigeon's egg, and even a hen's egg were not rare. The following
day some were picked up six centimetres across their widest dimension, there were even

some nearly one decimetre During the fall, there were several flashes of lightning

per second, the sky seemed set on fire." (Bibl. univ. de Geneve, t. LIX, page 339. 1887.)

(2) Annnaire de la Societe meteorologique, p. 129. 1865.

(3) Bulletin de 1'Association scientifique de France, p. 49. October 3ist, 1875.

181 %

of water vapoury for we know that it is the characteristic of rapid
crystalizations to lead to distorted and not transparent crystals.

This first core being formed, the spiral motion in the body of
cloudy moisture produces all round it a layer of ice, more slowly
formed and, consequently, transparent. After another electrical
discharge a fresh emission of vapour takes place, and, at the same
time, new hail stones appear ; those which are still turning may be
again covered with a second layer of vapour which is quickly
frozen into a snowy condition, and so on.(l)

217. As regards the origin of the formation of these whirl-
winds of hail, the experiments above described under the name
of electro-dynamic whirls (158) (Figs. 65 and 66), and which we
shall refer to further on (221), in order to explain the spiral motion
of waterspouts and cyclones, have led us to attribute it to the
magnetic influence of the earth. The appearance of hail is, in
fact, connected, as we have just seen, with the presence of great
quantities of electricity in the clouds, the discharges of which
constitute veritable electric currents, of short duration, it is true,
or rather intermittant, but possessing all the characteristics of a
powerful current of dynamic electricity.

Electric currents may turn under the influence of magnetic
action, and with so much the greater rapidity as they pass through
more mobile conductors. When these currents radiate in all
directions at the centre of a liquid, as in the experiment in
question (158), the spiral motion produced under the influence of

(i) However, one may demur by saying that the R. P. Sanno-Solaro has artificially
obtained small masses of frozen water all at once, without the successive addition of fresh
quantities of water, and still shewing concentric layers alternately opaque and transparent.

It would appear from this that the whirling movement is not absolutely necessary in
order to account for this kind of formation of hail stones. But in that case their formation
may be simply referred to the preceding explanation/which we have already set forth (215).

182

a magnet, and made visible by a cloud of semi-metallic powder
detached from the electrode, works in a spiral form with an extra-
ordinary rapidity.

Columns of cloudy or moist atmosphere, powerfully electrified
and changeable in all directions, should then likewise take a rapid
spiral motion under the influence of the earth's magnetism, and
so raising into whirlwinds the hail which accompanies the
discharges.

218. Thus, electricity appears to be connected with the pro-
duction of hail by the variety of its effects, either mechanical,
calorific, or magneto-dynamic.

The part played by the winds is, no doubt, important; they
entrain, divide, or re-unite in their passage the electrified cloudy
masses ; they bring together those strongly charged with electricity
with those less charged ; they raise them towards the cold regions
of the atmosphere, or facilitate the fall of temperature around
them necessary to solidification ; they also direct them, according
to the configuration of the ground, towards the points where we
notice that hail seems to appear by preference.

But these are the accompanying causes which only prepare
conditions favourable to' the production of hail, whilst electricity
is, in our opinion, the actual cause which, by its presence in the
clouds, and the instantaneous power of its discharges, causes the
sudden formation and fall of this phenomenon/1'

(i) Comptes rendus, t. LXXXII, p. 316, January 3151, 1876.

CHAPTER III.

Comparisons with Waterspouts. — Upon
the formation of Waterspouts and
Cyclones.

219. The effects obtained by means of high tension currents
of electricity, which have been described above (146 to 151),
present a great similarity to those of water-spouts, and shew the
importance of the part that must be played by electricity in these
great natural phenomena.

The experiment shewn (fig. 52, page 129), in which a liquid
vein or column, powerfully electrified, flows above a magnet,
reproduces the principal effects of water-spouts, in infinitely
reduced proportions, it is true, but with their most characteristic
features ; — the noise they make, the mist formed around them, the
luminous traits or silent lightning with which they are streaked,
the fiery globes which sometimes appear at their extremity, (I) and
the boiling of the water when they touch the surface of the sea ;

(i) We may, in fact, notice, in this experiment, at the point where the column meets the
surface of the liquid and all round the spark, small luminous liquid globules at the same
time that the steam and finely divided water appear.

184

so that these phenomena may be compared with positive electrodes
of liquid or vapour, from which powerful 'electric currents from
storm clouds escape toward the earth or the sea.

220. This experiment also leads us to attribute the spiral
motion of waterspouts'1' to the flow of the electric current under
the influence of the earth's magnetism ; for this motion takes
place in precisely the same manner as in the experiment in
question ; that is to say in the opposite direction to the hands of
a watch, supposing the observer to be placed in the northern
hemisphere, and in the same direction if the observer be placed
in the southern hemisphere (146).

221. If it be considered that this direction is the same as that
taken by cyclones, that the rotation of these great atmospheric
currents acts in a spiral form, according to the observations made
by numerous sailors, (2) like the electro dynamic spiral motions
already observed when an electric current, debouching at a certain
point, radiates in all directions over a magnet (158 to 159); if it be
also remarked that these spiral motions are accompanied by the
most intense electrical manifestations at their starting point in the
tropical regions, (3) and that cyclones seem to be developed round
a point called the "eye of the cyclone," which is a veritable

(1) Most Waterspouts, either on earth or sea, are accompanied, as we know, with a
spiral motion. At the approach of a water-spout "the surface of the sea begins to be
disturbed ; we see the water foam and turn slowly, until the rotatory motion becomes
accelerated." . . . (Dampier, Voyage au tour du Monde.)

(2) V. Notes sur la forme des cyclones dans 1 'ocean Indien, etc., par Meldrum, directeur
de 1'observatoire de Maurice. According to Mr. Wilson, manager of the meteorological
Observatory at Calcutta, the form of cyclones in the Gulf of Bengal is also rather spiral
than circular.

(3) See the works of Reid, Piddington, de la Poey, de la Havane (Tempetes electriques.
Annuaire de la Societe meteorologique, 1855) cle Marie-Davy (Les mouvements de 1'atmos-
phere et des mers), de A. Le Gras, Bridet, Roux, Zurcher et Margolle(Trombes et cyclones),

185

furnace of electricity, we may be permitted, we think, to attribute
these formidable phenomena to the magneto-dynamic rotation of
atmospherical electric currents, to which the clouds act as
moveable conductors, and the movement of which is communi-
cated to the volume of the air surrounding them.(l)

222. We will add in support of these assumptions, as we have
already done with regard to whirlwinds of hail, which have, we
believe, the same origin (217), that the speed with which these
electro-dynamic motions are produced in our experiments, is very
great, — the electric current not being confined to metallic con-
ductors, but being free to expand from a single point in all
directions through the body of the liquid.

Looking at the rapidity of these motions, we can conceive the
power that these atmospherical currents may acquire when free to
move in every direction, and charged with a great quantity of
electricity radiating equally in all directions in the body of the
atmosphere, and transformed by the earth's magnetism into spiral
motions/2'

223. The spiral motion of simoons, as frequently seen in
India, the electric origin of which has been demonstrated by

(1) Although the spiral motion of cyclones may have been generaly attributed, by most
of the writers upon this question, to the meeting of winds of contrary direction, or of
different rates of progress, we will add, however, that Reid himself, one of the authors
of the Lois des Tempetes, assumed that " electro-magnetism had perhaps some connection
with the rotary character of storms, and their opposite motions in the different hemispheres."
"(H. Piddington, Guide du Marinsur la lot des tctnpetes, p. 171, Paris, 1859) • • • • but
all this is so speculative," adds Piddingtont " that we will limit ourselves to merely pointing
it out."

(2) We will not here consider, be it understood, the motion of translation or trajectory
of cyclones, which depends upon the direction of the higher and regular winds, combined
with the earth's motion of rotation.

186

Dr. Baddeley, may likewise be explained by the influence of the
earth's magnetism. (I)

224. The experiment above described (146), further proves
that waterspouts, even when they are not accompanied by any
sign of electricity, may, nevertheless, be electrically charged, and
owe their spiral motion to the actual presence of this electricity.
The fact is, they form, in this case, a conductor sufficiently perfect
for the electric current to flow through it, without being transformed
into heat or light.

225. The same experiment also proves that waterspouts must
be charged with positive electricity ; for if they were negative, the
spiral motion would take place in the opposite direction to that
now observed in either hemisphere.

226. The formation of waterspouts, or the descent of these
cloudy appendages towards the earth, has been attributed, with
some reason, by Brisscm(2) and Peltier, (3) to an electrostatic attrac-
tion between the clouds and the earth. To this very naturally
attractive force may be added a carrying action, like many

(i) The lighting itself sometimes presents a spiral form. Coulvier-Gravier (Recherches
sur les meteores), M. de Fonveille (Eclairs et tonnerrej, and other observers have quoted
various examples.

This kind of lightning arises, according to all probability, from the same cause, and seems
to us to be comparable with a neat electro-dynamic experiment, due to M. Le Roux
(Annales de Chimie et de Physique, 36 serie, t. LIX, p. 409, 1860 ;— et Traite de Physique,
par Daguin, 36 edit., t. III. p. 649), in which a very fine and very flexible wire, suspended
vertically, raised to a red heat by an electric current passing through it, becomes
spontaneously curled in a spiral form, with great rapidity, round a magnet.

The luminous electric flow of the lightning must also turn under the magnetic influence
of the earth, and, assuming that this is a descending current of positive electricity, the
course described would be a right hand spiral in the northern hemisphere, and left hand in
the southern hemisphere.

" See for more detail upon this subject a note published in La Nature, April 7th, 1877,
p. 3°o-"

(2) Brisson. Traite de Physique, t. Ill, p. 418, Paris 1803.

(3) Observations et Recherches Experimentales sur les causes qui concourent a la
formation des trombes, par Ath. Peltier, Paris, 1840.

187

examples afforded by dynamic electricity, which tends to
facilitate the flow of water from the electrified cloud. It can
easily be conceived that a very dense mass of clouds, powerfully
charged with electricity, might give rise to the fall of a watery
column when it passes sufficiently near the earth or the sea.

227. The tidal wave, which often accompanies cyclones/1' the
"bores" (seiches) of the Swiss lakes, consisting in a sudden rise of
the waters in the form of waves or undulations, especially at the
narrowings of the lakes, which are particularly produced during
violent storms,'2' may be also explained by electric actions, as
Bertrand and other observers have thought.

The experiment we described under the name of " electric bore"
(147), figs. 53 and 54, in which a high tension electric current
gives rise to one or more small waves upon the borders of the
surface of the liquid considerably raised above its ordinary level,
shews that a current of atmospheric electricity may depress or
elevate liquid masses, just like a breath or strong wind, and goes
further to prove the electric origin of these phenomena.

228. The phenomenon of the ascension of a liquid column,
produced by the actual flow of a powerful electric current, which
we described above under the name of voltaic pump (148), fig. 55,
the aqueous cones formed below an electrode which conducts the
current to the surface of the liquid (151) allows the explanation

(1) This coincidence of the tidal wave with the cyclone is very remarkable ; there is no
instance of such a storm having struck La Reunion without being preceded by a phenomenon
of this kind." (J. Rambosson, Histoire des meteores, p. 243).

Piddington quotes a case observed at Ramsgate, "from the sudden rise and fall of the
tidal wave at this port in August, 1846 ; they occurred three at a time by unequal undulations
during a heavy storm and exactly at the same time as a strong discharge of what is called
the return shock of the electric fluid." (Guide du marin, p. 136).

(2) See the recent works upon this subject by M. Forel, Bibl: univ. de Geneve, aout et
September, 1878.

188

of the effects of very strong suction produced by waterspouts, and
of conceiving especially how, in waterspouts of tubular appearance,
this suction, acting throughout the whole length of an electrified
column, can raise water to an indefinite height, a fact which has
caused these phenomena to be called by the name of "pumps" or
"syphons" in certain parts of the world. The water sucked up
may arise from the sides of the vapourous channel itself, and thus
is explained the fact of the absence of salt in the water falling
back from these marine waterspouts.

229. The phenomena produced by static electricity also shew,
in a more feeble degree, effects of suction and evaporation, of
which Brisson and Peltier have already pointed out the similarity
to waterspouts. But these that we have observed with strong
currents of dynamic electricity, combining at the same time both
quantity and tension, appear to admit of still closer comparison
with the conditions of nature, and we think we may conclude
from this experimental study, that waterspouts and cyclones are
powerful electro-dynamic effects produced by the combined forces
of atmospheric electricity and the earth's magnetism.'1'

(i) Comptes rendus, t. LXXXI, pp. 185 et 616. July 26th and October nth, 1875; —
et t. LXXXI I, p. 220. January i/th, 1876.

CHAPTER IV.

Comparisons with the Polar Aurora. -
Upon the formation of the Aurora,
and the origin of Atmospheric
Electricity.

230. The well known experiment of De la Rive upon the
rotation of electric gleams, produced in a vacuum round a magnet'1'
has already suggested the electric origin of the polar aurora, and
its connection with the earth's magnetism.

But a certain number of circumstances which accompany their
appearance still remain to be explained.

The experiments above described (157), figs. 61 to 63, in which
the electric current comes into the presence of watery substances
or moist surfaces, as in the atmosphere, appeared to us to present
a series of phenomena exactly similar to those of the aurora.(2)

231. In fact, we recognise here, in spite of the diminutive
proportions, crowns, luminous arcs with brilliant fringes of rays,
regular or twisted, and possessed of a rapid undulating motion.

(1) V. Traite d'electricite, par A. de la Rive, t. II, p. 248, et t. Ill, p. 289.

(2) Comptes rendus, t. LXXXI, p. 185, July 26th, 1875; et t, LXXXII, p. 626, March
ijth, 1876.

190

This undulating motion, especially, offers a perfect analogy with
that compared in the aurora to the undulating movements of a
serpent, or to the folds of a cloth shaken by the wind.

232. Although the yellow light prevails in these experiments,
on account of the salt solution used, there may also be observed
at points where the water arising from the condensed steam is less
charged with salt, purple and violet tints similar to those of
the aurora.(I)

233. The rays from the luminous arc of the aurora should
arise from the penetration of the electric current into the moist or
frozen masses with which it meets, in the same way as those
observed in these experiments. The resulting vacuum being
gradually re-filled, fresh rays keep continually forming, and thus
is explained how the points of light from the aurora scintillate,
or appear shot forth and renewed every moment.

234. The dark circle or segment formed in the aurora by the
mist or cloudy veil, with which the electric current comes in
contact, corresponds in the experiment with the moist circle or
segment surrounding the electrode and round which the voltaic
current expands.

The portions in the immediate neighbourhood of the point
whence the electric current flows, being vaporized, it is only at a
certain distance that the electric wave is arrested and transformed
into heat and light.

235. The similarity of form between the luminous arc pro-
duced in our experiments and that of the aurora is also very

(i) See the descriptions of the polar aurora in the works of A. de Humboldt (Tableaux
de la Nature, Cosmos), Bravais, Lottin, &c. (Voyages et Scandinavie), Arago (Notices
Scientifiques), Piazzi-Smyth (Observations made at the Edinburgh Observatory, 1877, vol.
xiv, pi. 5, 6 et 7), Liais (Les Mondes, t. XXXVIII), etc.

191 •

striking. The current, passing in this manner through the volta-
meter, arises from the solution not entirely surrounding the
electrode. But if the electrode be more deeply immersed, waves
or complete luminous circles are produced (157), the same as in
the aurora, the arc of which is often considered as the portion
visible to an observer, of the complete luminous circle.

236. In the same experiments, we have noticed that the liquid
is violently disturbed by the electric current; whirlpools are
formed by the meeting of the electrified waves with each other,
and, if we only use a small quantity of liquid, a luminous boiling,
corresponding to that fluctuation in the light which also character-
ises the polar aurora, takes place.

237. As the electrode penetrates deeper into the liquid, the
water vapour is set free in proportionately greater abundance.
This phenomenon, of which the strongest batteries of static
electricity hardly allow us to form a suspicion, is an important
matter for consideration ; for it explains, in a natural manner, the
heavy falls of rain or snow which are always noticed during the
continuance of the aurora.(l)

238. The production of strong winds consequent upon the
appearance of the aurora-borealis, as we have already remarked
with respect to hail, shews that the discharge of a great quantity
of electricity in the atmosphere causes the formation of powerful
atmospheric currents by its calorific action, and the instantaneous
vaporization and rapid condensation resulting.

(i) " Often the aurora-borealis is accompanied with hoar frost, and the greater part of
them are followed by heavy falls of snow or rain, or by violent gusts of wind and storms."
(Extrait d'une communication de Meeker de Saussure & Arago. Notices scientifiques, t. I,
p. 694.)

192

239. The noise often heard during the aurora is due, like that
in our experiments, to the vaporization produced in the track of
the electric spark penetrating a moist substance.

"This noise is," they say, "especially intense when the
rays dart forth with rapidity."'1' The noise in the voltameter is
also so much the more intense in proportion to the length of the
rays fringing the luminous arc, and the rapidity with which they
are formed in the body of the liquid.

240. The magnetic disturbances caused by the aurora may be
re-produced in these experiments by placing a magnetized needle
close to the circuit. The deviation increases or diminishes
according as the luminous arc is more or less developed in the
liquid.

241. From these facts we may further conclude that the aurora
must be produced by a flow of positive electricity ; for the lumin-
ous phenomena are similar to those of the positive electrode in
the voltameter, and the negative electrode shews nothing like it.

242. But is the polar aurora caused by a discharge between
the positive electricity in the atmosphere and the supposed negative
electricity from the earth ? If that were the case, thunder should
be very frequent at the poles, or the gleams and luminous sprays
upon the projecting points of the earth, forming the counterpart
of the phenomenon passing in the air, would be observed. Now,

(i) Kaemtz. Traite de meteorologie ; traduction de Ch. Martins, p. 428.

The existence of such a noise has been doubted by some observers ; but the mass of
evidence from people inhabiting the region of the aurora-borealis goes to prove that this
noise is sometimes heard, no doubt when the height at which the aurora is produced is not
too great. (V. Arago. Notices scientifiques, t. I, p. 693.)

This is what Dr. Hjaltalin writes in a memoir upon the aurora-borealis : " I first directed
my attention to discover if any noise did or did not accompany the aurora-borealis ; I feel
sure that this noise does exist, although it is comparatively rarely heard, and I have only
heard it six times in a hundred observations." (V. 1'Annee scientifique, 1864, par Louis
Figuier, p. 107.)

193

actual observation shews that it is not so. We are then led to
think that it is the partial vacuum in the high regions which,
acting like an immense conducting envelope, plays the part of the
negative electrode in the experiments referred to above, and that
the positive electricity flows towards the planetary spaces, and not
towards the earth, across the mists or frozen clouds which float
above the poles.

243. Regarding the origin of this polar electricity, it was
admitted that it came from the equator and tropical regions. But
it may be objected that the electrified clouds must become
discharged during so long a passage, and, in fact, we know that
storms are more and more rare as the poles are approached.
Some analogies deduced from our experiments, which we shall
find further on (Chap. VI.), having led us to consider the heavenly
bodies as charged with positive electricity, — the only kind of
electricity which, perhaps, does exist, — we might be inclined to
regard the earth itself as charged with positive electricity, liberated
from the ground and the sea by means of simple emission, and
radiating from the whole surface at the poles as at the equator,
producing very different effects in the atmosphere, in consequence
of the entirely opposite meteorological conditions prevailing in
these regions.

244. The positive electricity thus arising from the earth would
not appear to be, in our opinion, the result of production or gener-
ation, properly so called, by physical or chemical causes. It could
not be due to evaporation, nor to friction, nor to thermo-electric
action, but should arise from a primary charge or store of electricity
proper to the earth itself, carried by it from the origin of its
formation and which would tend to be wasted, in the same way
as the heat it possesses, with extreme slowness, on account of
the large mass.

194

245. This electricity, penetrating the atmosphere, would con-
secutively attain the higher strata in which the air, becoming more
and more rarefied, affords immense conducting spaces, and would
spread from thence into the planetary regions. The lower beds
of the atmosphere, near to the earth, not being rarefied, we may
conceive that the positive electricity may only appear in a very
small quantity and accumulate in the greater altitudes. Thus
might be explained the increase in quantity of positive electricity
in proportion as we rise in the atmosphere.

246. The earth would not then act, in our opinion, as opposed
to the atmosphere, like a body which produces electricity by
the friction of another body taking up the contrary electricity.
Because we should then be led to admit that this production
takes place at the point of separation between the earth and the
atmosphere, without any apparent physical or mechanical action
other than the slow evaporation at the surface of the sea. Now,
it is known at the present time that the phenomenon of evaporation
is not in itself a source of electricity.

Vapour formed above the sea appears to us to merely constitute
a prolonged conductor in the atmosphere of the liquid conductive
mass of the earth, from which positive electricity emanates in
consequence of its high tension. It may thus be explained how
the liberation of electricity may be greater over the sea than over
the solid crust of the earth.

Water vapour facilitates the diffusion of electricity in the air,
which, at its ordinary pressure, acts like an insulating body.

247. We may also comprehend, according to this hypothesis
of the earth considered as an electrified body throughout its

195

entirety, that this electricity itself might be set free by means of
eruption, and form volcanic storms, always accompanied by light-
ning and thunder, and happening simultaneously with earthquakes,
which are connected with the internal movements of a liquid
melted mass, charged with electricity. The vapour produced
above this melted mass, finding an outlet by the volcanoes, must
necessarily carry away electricity, the same as the vapour formed
above the sea, without there being any occasion to admit chemical
subterranean effects producing electricity by means unknown to us.

248. If we now consider this emission of electricity in
equatorial and tropical regions, where evaporation is very abundant
there would naturally result clouds strongly electrified, and
continual storms.

These clouds could not directly become raised to a great
height ; because they are carried away by the winds prevailing in
these regions, and the electrical phenomena, appearing above the
points where the storms have originated, continue to be produced
in their passage, but become weaker as the latitude increases.

249. At the poles, on the contrary, where evaporation is far
less rapid, the quantity of electricity tending to emanate from the
earth is doubtless much less abundant, for the earth, being drier
at the surface of the ground or the sea in these regions, becomes
less easily charged ; but that which is liberated rises directly into
the higher strata of the atmosphere, and thus forms a kind of
electrified curtain, tending to become diffused in space.

If no moist conducting mass happens to be interposed between
this flow of electricity from the already elevated parts of the
atmosphere towards still greater altitudes, the electricity becomes
discharged in an invisible, or only slightly luminous, manner ; for

196

the transition from the less rarefied portions of the air to those
more rarefied is not quick but gradual. Its passage does not then
make itself known except by magnetic disturbances.

If, on the contrary, masses of clouds, in a condition of liquid
globules or crystals of ice, float in the intermediate space, lumin-
ous effects are shewn, as in our experiments, and we see the
polar aurora.(l)

250. This way of looking at the earth, as charged with positive
electricity as well as the atmosphere itself, seems to render
inexplicable, at first sight, the discharges which take place in
ordinary storms between the clouds positively electrified and the
equally positively electrified earth.

But this apparent difficulty is easily resolved if we consider that
a given portion of the surface of the earth, although emitting
positive electricity, is much less charged with it than the cloudy
mass which passes above, after having gathered and stored in its
course the positive electricity spread throughout the atmosphere,
and also bringing a large part of that taken, during the formation
of the cloud itself, above the sea in the warm regions.

The consequence is that this part of the ground, having only a
comparitively weak positive tendency, becomes strongly negative
by induction. It is the same as between the clouds themselves,
which may be all positively electrified and yet be the seat of
powerful discharges, because they are electrified in different
degrees.(a)

(1) " The more recent observers place the seat of these phenomena, not at the limit of
our atmosphere, but in the region where clouds and vesicular vapour are formed." (V.
Cosmos, par A. de Humboldt ; traduction de MM. Faye et Galuski, 46 edit., t. I, p. 224).

(2) Messieurs Quetelet and Palmieri have allowed, as is known, that clouds which appear
negative are only so by induction, and only at one of their extremities.

197

251. The arguments we have just set forth agree also, up to a
certain point, with Ampere's hypothesis, admitting the existence
of an electric current in a fixed direction, encircling the earth and
producing its magnetic action.

We need only add, in order to explain the emission of electricity
into the atmosphere, and, consequently, into space, that it must be
a current of very high tension, impossible to be confined in a
conductor of limited material like the wire of a solenoid, but
radiating all round the mass of the earth on account of its high
tension.(l)

252. Finally, we draw the conclusion from these premises, that
the polar aurora results, in our opinion, from the diffusion of
positive electricity, emanating from the polar regions themselves,
in the higher strata of the atmosphere round the magnetic poles,
either in invisible rays when there are no clouds interposed, or
converted into heat and light by meeting with aqueous substances
in either a liquid or solid state, which it vaporizes with a noise,
and precipitates in the form of rain or snow upon the surface of
the earth.<2>

(1) This paragraph and those which precede it from 244, respecting the origin of atmos-
pheric electricity, are extracts from an article presented to the Academic des Sciences on the
ijth March, 1876, from which an extract only was inserted in the Comptes rendus.

(2) Comptes rendus, t. LXXXII, p. 629, March i8th, 1876.

CHAPTER V.
Comparisons with Spiral Nebulae,

• 253. If spectral analysis has permitted, in these later times, of
the study of the chemical composition of celestial bodies, it is not
chimerical at the present time to try to account for their physical
constitution by the observation of the electrical phenomena and
by the inferences to which these phenomena lead.

Herschell and Ampere had already thought that the incandes-
cence of the sun might be attributed to electric currents. Several
astronomers and modern scientists, among whom we may mention
Messrs. Young, Morton, Respighi, Spoerer, Marco of Turin, have
suggested similar ideas.

The phenomena we have observed with very high tension
electric currents, such as the spiral motions, the luminous effects,
the spherical or annular form taken by substances submitted to
the action of the current, have led us also to think that electricity
in a dynamic condition might play an important part, not only in
meteorological phenomena, but also in those of celestial science.

The experiment described above (158), in which a cloud of
metallic oxide, torn from an electrode by the current, takes a spiral

199

motion in the body of a liquid under the influence of a magnet,
seemed of a nature to explain, in particular, the remarkable form
of spiral nebulae.

It is, in fact, sufficient to take a glance at the figures shewing
this experiment (figs. 65 and 66, page 142), in order to at once
recognise the form of these nebulae described by Lord Ross, some
of which have the curve of their arms pointed in the opposite
direction to the hands of a watch, as in fig. 65, such as the nebula
of the virgin ("Chevelure de Berenice"), &c.; others have
their spiral turns directed the same way as the hands of a watch, such
as those in fig. 66, and like the nebula "Chiens de chasse," etc.(l)

254. In view of so striking a similarity, may it not be reason-
ably supposed that the nucleus of these nebulae may be formed
by a veritable electrical furnace ; that their spiral form is probably
caused by the presence of celestial bodies powerfully magnetised,
and that the direction of the curve of the turns in the spiral must
depend upon the nature of the magnetic pole turned towards the
nebula(2)?

It would then be interesting to search, among the stars already
known round these nebulae, those which, by their position, could
exercise this magnetic influence, or to explore the celestial
firmament, on the axis of which the spirals appear to turn, within

(1) V. Le Ciel, par. Araedee Guillemin. 56 edit., p. 833 et suiv.

(2) These inferences may be objected to on the score that no conductor can be perceived
in space leading an exterior electric current to the centre of the nebulae. In answer to this
objection we will call to mind that, among other experiments made with a far weaker source
of electricity, we have noticed small luminous rings composed of incandescent particles
entirely detached from the electrode ; these rings, the centre of which is disturbed by a
whirlpool, are set in motion in the space comprised between the electrode and the larger
luminous ring formed around by the shock of the electric wave against the sides of the volta-
meter (157). They are in this case veritable electric fires, separated from the principal source
which gave rise to them, and similar to, although infinitely smaller than, the nucleus of
isolated stars, such as those which form annular nebulae.

200

or beyond the plan according to which they become developed, in
order to discover celestial bodies capable of causing their form or
spiral motions.'1'

In the case where a star might be known to fulfil these
conditions, we might still examine, along the line passing through
the centre of the nebula and the star itself, if there were not a
second spiral nebula connected with the other magnetic pole of
this star, the curves of which, turning in the opposite direction to
the magnetic currents of this pole, might, nevertheless, appear to
the observer to be directed in the same direction as those of the
first, and the combination of these three bodies would thus form
a symetrical Stella system.

Cosmic matter is spread with such profusion throughout space
that this hypothesis partakes in no way of the impossible.'2'

As such research requires the use of the most powerful
telescopes, we only beg to draw the attention of astronomers
thereto with every reserve which inductions based upon simple
analogies demand ; but, if actual observation happened to justify
them, it would assuredly be a decisive proof in favour of the
electrical constitution of the celestial bodies.'31

(1) Reference may also be made to electro-dynamic actions of this kind in order to
explain the breaking up of celestial bodies as supposed by Mr. D. Vaughan of Cincinnati
(July, 1878). This breaking up may also be facilitated, or begin, by the cooling of the stars
being carried to its utmost limit, as supposed by M. Stanislas Meunier, in order to explain
the origin of meteors. (Le Ciel geologique, p. 195, 1871).

(2) The nebula of the "Chiens de Chasse" itself affords another nebula which has been
recognised by Chacornac as aho shewing a spiral form.

(3) Comptes rendus, t. LXXXI, p. 749, October 26th, 1875.

CHAPTER VI.

Analogy with the Solar Spots. — Upon
the physical formation of the Sun.

255. The electrical perforations made by high tension currents
(160), figs. 68 to 70, appeared to us to present some remarkable
similarity to the formation of the solar spots, such as those
observed by Messrs. Nasmyth, Dawes, Lockyer, Chacornac,
le P. Secchi, Tacchini, Langley, etc., and which have been
compared with sprigs or bundles of thatch, and filaments, crooked,
twisted, or entwined, etc.

The material in action in our experiments is very different, it is
true. In this case it is simply moist organic matter, whereas, in
the sun, it is a question, no doubt, of a fluid and incandescent
mineral substance. But the action of electricity might take
place, and shew itself in the same manner, in both cases : in the
first, by sub-division of the dried up matter and jets of water
vapour into innumerable threads ; in the second case, by extremely
finely divided streams of luminous matter, and by ejections of the
vapour from mineral substances.

This curious appearance of spots on the sun, so difficult to
account for by ordinary mechanical action, is easily explained
by the intervention of electricity ; one of the most characteristic

202

properties of which is to form into points, or to cut up into
threads, all matter opposing its passage in order to open up the
many tracks which seem necessary for its rapid discharge.

256. We have thus been led to think that the spots upon the
sun are cavities produced by essentially electrical eruptions;
that, in consequence, the interior mass of the sun must be strongly
charged with electricity; and that, according to the direction of
the excavations, the thread-like slope of which turns towards the
interior of the star, the electricity which escapes must be
positive (160).

257. But these are not the only analogies which we can refer
to in order to prove the presence of electricity in the sun. The
phenomena observed in incandescent metallic globules, formed
under the action of a powerful electric current, may also throw
some light upon the physical formation of the sun, considered
simply as a globe of incandescent matter. (l)

(i) Le Verrier wrote in 1860, after having observed a total eclipse of the sun in Spain : —
" I have got at the physical formation of the sun. It seems to me that we must completely
abandon the present theory. The thing may be much simplified."

We used to be told that the sun was composed of a dark central globe ; that above this
globe there was a deep atmosphere of dark clouds ; higher still there was supposed to be
phosphorus in the form of a gaseous envelope, luminous in itself, and the source of the
light and heat of the sun. When this substance became torn, it was said that the dark
centre of the sun could be perceived, hence the spots which frequently show themselves.
To this complex formation it was necessary to add a third envelope, formed of a combination
of roseate clouds.

Now, I fear that the greater part of these envelopes are the work of imagination, that
the sun is simply a luminous body by reason of its high temperature, and covered by a
continuous layer of the roseate matter, the existence of which we are at present aware. The
star, thus formed of a central liquid or solid body, covered by an atmosphere, follows the
law common to the constitution of celestial bodies.

Whatever the formation of the heart of the sun may be, solid or liquid, the surface and
the interior of the star must be at least as much disturbed as the surface and interior of the
earth, and neither waterspouts, electrical phenomena, nor volcanoes should be wanting,
capable of producing the movements observed." (Bulletin de 1'Association Scientifique de
France, p. 97, 1869).

203

We have, in fact, seen that these melted globules (fig. 13 and 14,
page 49), shew brilliant eruptions in consequence of the disturb-
ance in their interior substance under both the calorific and
chemical action of electricity, that the jets of gas and incandescent
particles arising from these interruptions necessarily made their
way out by the cavities or perforations produced in the interior of
the globule itself; and that these perforations, allowing the
comparitively colder and less luminous interior of the globule to
be seen, formed dark spots upon its brilliant and undulating
surface, and that, in consequence of the greater or less thinness of
the melted envelope around the craters, some parts of the surface
appeared more or less brilliant in the neighbourhood of the spots;
that these globules, examined after cooling, shewed a wrinkled
and pimpled surface ; finally, that they were hollow and that their
envelope was proportionately thinner as the metal enclosed more
gas in combination.

258. We have then arrived at the following conclusions from
these experiments by way of comparison : —

First, that the sun may be considered as a hollow electrified
globe, full of gas and vapour, covered with an envelope of melted
incandescent matter.'1'

Secondly, that the wrinkles or lucuke on its surface arise
from the undulations of this liquified envelope.

Thirdly, that the spots are simply perforations in the fluid
envelope, produced by quantities of gas and electrified vapour,
emanating from the interior of the star, and giving to the sides of

(i) This conclusion agrees with the known low density of the sun.

204

the cavities, as above described, the formation which characterises
the passage of positive electricity.

Fourthly, that the faculse appear to be a brilliant phase in the
evolution of the gaseous matters, when they approach the surface
previous to their eruption.

Fifthly, that the protuberances are formed by the gases them-
selves escaping from the interior of the star at a higher temperature,
and, consequently, more luminous than those forming the atmos-
phere upon its surface.

259. It may be objected that the metallic globules in question
are produced between the two poles of an apparatus and traversed
by an electric current, while the sun is isolated in space; but
referring to our former experiments, such as that of the spray
(143), one can understand the formation of electrified spheroids,
totally detached from the source whence they arose, and carrying
with them a considerable quantity of high tension electricity.
Besides, if, in the experiment of the metallic globule, we allow the
wire, to which the globule adheres, to melt, the current is broken ;
the globule remains suspended from one of the poles, and, during
the brief moment that it remains incandescent, spots are still pro-
duced, and bubbles are liberated upon its surface (fig. 14, page 49).

If this phenomenon lasts an appreciable length of time with so
small a mass it may be understood what duration it might possess
when the gigantic globe of the sun is in question. The vibratory
electric motion communicated to it must last like mechanical
motion as long as its own physical and chemical effects.

Thus, we believe that the sun is electrified, but that it does not
create the electricity which it possesses, any more than the heat

205

and light which arise from it; it is a store received from the
nebulous ring, of which it is but a brilliant particle, condemned to
become extinct some day ; this nebulous ring would be derived
from another electrified wave and so on up to the primary cause,
the creator of all power and motion. Taken from this point of
view, the incandescence of the solar globe, prolonged through a
series of ages, would be in itself but a spark of short duration in
the infinitude of time and space.'1'

(i) Comptes rendus, t. LXXXII, p. 816, April loth, 1876.

FIFTR PART.

Rheostatic Machine.

260. Description of the rheostatic machine.— After

having described apparatus suitable for accumulating and trans-
forming the work of the voltaic battery, so as to obtain, at will,
temporary effects of quantity or tension very much higher than
those of a given battery, and after having specially applied them to
the observation of effects produced by electric currents of high
tension, and having seen how much interest the study of these
effects created by their analogy with natural phenomena, we have
sought to more completely transform the energy of the voltaic
battery, and to obtain an E.M.F. equal to that of electrical
machines.

This problem seemed already solved, no doubt, by induction
apparatus; but the method we used, although less simple from
a practical point of view, appeared to us more direct in theory,
and capable of converting, with less loss in transformation, a given
quantity of electric energy ready to provide a dynamic current,
corresponding in quantity with electrical effects of the static kind.

207

We have already had occasion to frequently prove that our
secondary batteries of 600 to 800 elements permitted of the rapid
charge of a condenser having a sufficiently thin insulating plate
of glass, mica, gutta-percha, paraffin, &c.(l)

Fig. 74.

In order to obtain continuous static effects of the greatest
intensity, we joined a certain number of condensers, formed, by
preference, of mica coated with tin foil, arranged like the secondary
batteries themselves, so as to be successively charged in parallel
and discharged in series; and we called the apparatus thus invent-
ed, by the name of rheostatic machine, (fig. 74).|2)

261. All the plates of this apparatus must be insulated with
great care. The commutator is formed of a long vulcanite
cylinder furnished with longitudinal metal strips intended to
connect the condensers in parallel, and pierced transversely by

(1) We know that Volta, Ritter, Cruikshank, etc., were able to charge condensers by
means of the battery, and that these results gave occasion for subsequent research, by
calculation or experiments, on the part of many scientists.

(2) Comptes rendus, t, LXXXV, p. 794, October agth, 1877.

208

copper wires bent at each end, or with metal hooks slightly
rounded, for the purpose of uniting the condensers in series.
Wires in the shape of springs rr are connected with the two
armatures of each condenser and fixed upon a bar of ebonite on
either side of the cylinder which may be rotated rapidly by means
of the wheel R and a pinion geared with it.

When the cylinder is turned so as to present its longitudinal
metal strips in contact with the springs, the even row of all the
condenser armatures are united on one side and the odd row are
united on the other side, so as to form a single condenser of large
surface, charged by connecting the terminals P and P' to the poles
of the battery.

When the cylinder is turned the other way, as shown in the
figure, so as to present the transverse wires to the springs, all the
charged condensers become united in series, or tension. The
armature L of the condenser on the extreme left communicates
with the last spring on the far side of the cylinder, and terminates
at the arm E of the exciter. The armature L' of the last con-
denser on the right communicates with the last spring but one on
the next side ; this spring comes in contact with the last metal pin
which pierces the cylinder from one side to the other, and the last
spring on the far side of the cylinder communicates with the other
arm E of the exciter.

Whilst the condensers are thus united in series, the battery
charging the apparatus is entirely thrown out of the circuit ; the
last spring /, seen in the figure, communicating with the battery
through the terminal P', and the other end spring communicating
with the terminal P, do not touch any metallic portion of the
commutator cylinder.

209

262. Effects produced by the rheostatic machine.*1'—

To study the effects produced, we have used several machines,
composed of a varying number of condensers having a double
armature of about three square decimetres in surface.

By first employing a machine with six condensers, charged by
the secondary battery of 800 cells, we obtained, at a speed of
fifteen revolutions per second, a series of brilliant sparks, from
thirteen to fourteen millimetres in length, following one another at
the rate of thirty per second, and forming a continuous stream of fire,
accompanied with the same noise as that of the sparks from an

i**. «,«>>. ^eoci&w, w. j t.U

induction coil supplied by a Leyden jar.

With machines composed of thirty, forty and fifty condensers, we
have obtained sparks of four and five centimetres.

263. The length of the sparks increases nearly in direct ratio
with the number of condensers; but it is not possible to establish it
definitely on account of the inequality in the thickness
of the insulating plates and the variable effects resulting
from it. Thus, with a series of thirty selected condensers, with
very thin insulating plates, the sparks are four centimetres ; with
fifty condensers of various thickness, they hardly exceed five
centimetres.

264. On the other hand, when the insulating plate of the con-
densers is too thin, it is pierced by the current, and we then obtain,
if the current continues to act, the wandering spark described
above (fig. 48), p. 125).

265. The diameter of the commutator cylinder must be
proportioned with regard to the length of the sparks expected to

(i) Comptes rendus, p. 761, March 25th, 1878.

210

be obtained. One quarter of its circumference, or the interval
comprised between the metal strips of the commutator and the
line of transverse pins, must be greater than the length of the
spark which can be created ; otherwise the latter would strike
across the commutator instead of appearing between the arms
of the exciter.

266. Crooked and " brush" form of the discharge.-

The spark created by the rheostatic machine when a constant
and sufficient speed of rotation, and E.M.F. of the primary current,
is used, presents a special and very regular form that is not seen
with the same degree of clearness in that from electrical machines
or induction coils.

Fig. 75-

This form, when the angle comprised by the arms of the exciter
is very obtuse, consists in a stream of fire starting in a line with
the positive arm, and rising noticeably above the negative point, to
which it returns in the form of a crook, making numerous deviations
in descending upon this point (fig. 75).

When the distance between the points is increased by one or
two millimetres, the discharge given by the machine takes the form

211 .

of a brush, following the same course as the former one. A
luminous spray springs forth from the positive pole for about three
quarters of the distance between the poles and turns back towards
the short brush formed round the negative point (fig. 75).

267. The difference in the form of these sparks and "brushes,"
compared with those from induction coils, is accounted for by the
rheostatic machine not giving, like these latter apparatus, an alter-
nating flow of electricity, but always in the same direction, which
allows, besides, of its easy measurement with Thomson's long scale
electrometer for tension, and of its comparison with that of
electrical machines.

268. There is also occasion to remark, as we have above fore-
seen (260), that the loss of energy resulting from the transformation
of dynamic into static electricity is, in this case, much less than in
induction apparatus ; because, the voltaic circuit not being for a
moment short circuited, there is not that conversion of a part of
the current into heat. The current simply spreads over the
polarising surfaces presented by all the condensers in proportion
as they are discharged.

269. With machines of thirty to fifty condensers, giving
sparks from four to five centimetres long, and the speed of rotation
not being so great, the sparks are less continuous, and their form
is not so regular; their deviations arise or descend irregularly
above or below the straight line between the two points of the
exciter (fig. 76).

The positive "brush" then presents a calyx terminated by a
luminous ovoid sheaf, more or less branching out, just like that
from electrical machines (fig. 76).

212

270. Light produced in vacuum is brighter, in this case, than
from electrical machines, in consequence of the greater quantity
of electricity used, and, when the rotation is quick enough, it is
as bright and continuous as from induction coils.

Fig. 76.

Geissler tubes of the highest resistance, and M. Ed. Becquerel's
tubes of phosphorescent substances, are illuminated in a brilliant
manner.

The absence of all stratification, and of the sheath of blue light
round the negative pole presented by Ruhmkorff coils, is noticed.
The light is of a purple colour throughout the tubes, and
completely fills them in the same way as with an induction coil
and Leyden jar.

This effect must arise from an excess of tension ; for, if that of
the battery used to charge the rheostatic machine be much
reduced, the blue sheath and stratifications appear.

271. Light in a Vacuum.— The secondary battery of eight
hundred cells used to charge a rheostatic machine, although not
specially well insulated, is able to illumine Geissler tubes directly,
by producing in them stratifications such as those observed by
Messrs. Abria, Grove, Gassiot, Warren de la Rue and

213 *

H. W. Miiller. A long column of water being introduced in the
circuit, we can, with a single discharge of the battery, make a
Giessler tube luminous for more than three hours and a half, on
account of the very small quantity of electricity expended in the
passage of the current through the rarefied air.

But when the current from the battery is cut off whilst the tube
is being illuminated, it often happens that it is not reproduced by
reclosing the circuit. A slight diminution in the tension of the
battery is sufficient to prevent the phenomenon from taking place ;
because the E.M.F. of eight hundred secondary cells is about the
minimum tension that may be used to make electricity flow
through narrow tubes of rarefied air.

If we then introduce the rheostatic machine into the circuit, so
that the terminals of the vacuum tube communicate at the same
time with the poles of the machine and of the secondary battery,
we notice, in turning the machine for an instant only, the curious
result of the tube being immediately illuminated without stratific-
ation, and of the battery continuing to illuminate it alone with
stratification. The tube was "primed" by the E.M.F. of the
rheostatic machine being greater than that of the battery.

272. In general the rheostatic machine will give all the other
effects created by electrical machines and induction coils, and
these effects do not seem to be disturbed by variations in the
hygrometric condition of the air.

Production of the continuous spark or " brush " is accompanied
by a strong smell of ozone. Each of the poles will give sparks
when brought near to objects connected with the ground. The
electric "whirl" or "vane," or the effect of blowing produced
by the points of the exciter, are good examples.

214-

273. On the possibility of obtaining some effects

With a current Of less tension. — The apparatus in question
would offer but a theoretical interest, if it were always necessary
to employ a battery of eight hundred cells in order to show its
effects. We have consequently set to work to produce them with a
far less source of electricity and we attained the desired result by
increasing the number of condensers and decreasing as much as
possible the thickness of the insulating plates.

With a machine of fifty condensers with very thin mica plates,
held by frames of vulcanite, continuous sparks of six millimetres
were obtained by only using one hundred secondary cells and we
could even illuminate a vacuum tube by charging the machine
with a secondary battery of thirty to forty cells. (I) It is with this
relatively weak source that the blue sheath and striae are visible
round the negative pole.

274. Complete transformation of a certain quantity
of dynamic into static electricity. — It was interesting to
attempt to give an example of complete transformation, by means
of the rheostatic machine, of a certain quantity of dynamic
electricity stored by a secondary battery, and to know, approx-
imately, the time required to completely exhaust the charge in
static effects. Among the various experiments we have made may
be mentioned the following :

(i) In place of a secondary battery, a Bunsen battery of from fifty to sixty elements,
such as is set up for the electric light, or a Gramme machine wound for the highest possible
E.M.F. would do equally well for charging the rheostatic machine.

We have even obtained, with a very little secondary battery of six hundred elements
(formed of small forked plates of lead, of a few millimetres only in width) in connection
with the rheostatic machine, sparks almost as long as with six hundred of our ordinary
elements.

These six hundred little cells were immersed in glass tubes of one centimetre in
diameter, fixed upon a board only sixty millimetres long by forty millimetres wide. All
these cells were charged in series, by our large battery of eight hundred cells.

215 ,

A secondary battery of forty cells, without any residual charge,
but quite ready to store the smallest amount of chemical work of
the primary battery, was charged for fifteen seconds only by two
Bunsen elements and then coupled up to the rheostatic machine.
It was necessary to turn the apparatus for more than a quarter
of an hour to exhaust that charge in the illumination of a Geissler
tube (v. § 3000).

It may be inferred from this that, with the quantity of electricity
absorbed by a secondary battery in ten minutes (which is about
the best time to accumulate without noticeable loss the work of
the primary battery), we could maintain a vacuum tube luminous
for more than ten hours.

275. Length of the Sparks.— We have seen above (263),
that the length of the sparks produced by the rheostatic machine
was clearly in proportion to the number of condensers. But the
inequality in the thickness of the mica plates did not permit of
establishing this in a positive manner.

By employing, since that time, condensers with mica plates of as
uniform a thickness as possible, charged under the same conditions
by the secondary battery of eight hundred cells, and by only
making a half turn of the commutator of the machine, so as to
produce but isolated sparks, instead of a continuous stream of
fire, we have obtained some figures relatively higher; and, by
charging more regularly than the preceding, with a machine of ten
condensers, we have made sparks of a centimetre and a half; with
a machine of thirty condensers, sparks of four and a half centim-
etres and with a machine of eighty condensers, sparks of twelve
centimetres in length.'1'

(i) If we calculate the deviations described above, these sparks have a greater length ;
but we only in these cases measured the distance between the arms of the exciter.

WTBRSIT7,
" tf&

216

The length of the sparks produced by the rheostatic machine
may then be considered as proportionate to the number of
condensers.

276. This length increases more rapidly with the E.M.F. of
the current which acts upon the machine, and seems to vary in
proportion to the square of the number of elements, the same as
the direct spark from a battery of high tension, according to the
law given by Messrs. Warren de la Rue, and Hugo W. Miiller ;
but the results we have obtained were not always sufficiently
consistent to enable us to affirm, with certainty, that the spark from
the rheostatic machine exactly follows the same law.

277. Large rheostatie machine.— Figure 77 represents

the rheostatic machine of eighty condensers that we used in these
experiments. The vulcanite cylinder of the commutator is one
metre long by fifteen centimetres in diameter.(r)

The arrangement is otherwise nearly the same as that of the
first machine we described (261), but the end springs are
at a sufficiently great distance from those which are next to
them to prevent sparks striking across from the tension poles of
the rheostatic machine to those of the secondary battery. The
condensers are made with mica plates eighteen centimetres long
by fourteen wide coated on each side with tin-foil. Fine copper
wires covered with gutta-percha are fastened to the end of each
armature. The edges of the condensers are also fixed in frameSj
or only plain ebonite plates, in order to give them more

(i) According to the rule we have given (265), it appears impossible to obtain, with this
cylinder, sparks of a greater length than a quarter of the circumference (118 m.m.) ; but as
the points of the exciter, placed opposite one another, offer an easier means of discharge
than the space between the strips and pins on the commutator, the spark strikes across these
points even when the distance is a little greater.

217 .

rigidity and maintain them more easily in a vertical position close
to each other without contact.

218

278. Upon rotating the commutator, sparks appear upon all
the points where the metal strips meet the springs forming the
terminals of the condensers for charging them in parallel, and give
the cylinder the appearance of a sparkling tube.

Another time, sparks appear when all the condensers are
united in series, and a discharge is produced between the arms of
the exciter.

If a column of distilled water be placed in the circuit of the
secondary battery, the water seems to be decomposed in a
continuous manner whilst the machine is in motion. In reality,
this decomposition only takes place at the moment when the
sparks of charging are produced ; for, during the discharge, the
water tube, like the secondary battery, is thrown out of the circuit.

The limited quantity of dynamic electricity stored in the
secondary battery is thus expended little by little during the
charge only of the condensers ; but this expenditure is very slow,
and each charge of the condensers, and each discharge also,
corresponds with a very small quantity of electro-chemical
action consumed in the battery (300).

279. Sparks produced under different conditions.—

We have seen (275) that sparks produced through the air by
the rheostatic machine of 80 condensers, attained a length of
12 centimetres.

If flower of sulphur be spread between the two points of the
exciter, supported upon a plate of insulating material, the space is
rendered decidedly less in resistance and sparks may be thus
obtained 15 centimetres in length (fig. 77). If we cause them
to strike across a conducting powder, such as metal filings, they
will increase to 70 centimetres.

219

280. When these sparks traverse flower of sulphur they make,

in

Fig. ?#.
their passage, a sinuous furrow from 2 to 3 m.m. wide, and, if

220

the insulating surface, upon which the flower of sulphur is spread,
consists of a mixture of resin with about one tenth parafine, they
leave in the middle of the furrow a very clear bluish line visibly
traced, with a leadish appearance, which permits of the preservation
of the exact outline. (I) This trace is easily effaced, however,
by rubbing, but by carefully following it and scratching it by means
of a pointed instrument it is made ineffaceable and can then
be easily copied.

It is thus that we obtained fig. 78, which represents sparks of
various length the actual size.

281. Form Of the sparks.— We notice that these sparks,
when they have not attained their maximum length, often show
closed branches which might escape notice when the luminous
stream only is seen.

Their deviations are always rounded, and we never see that
zig-zag form with sharp angles in which the electric sparks are
often represented. The sinuous form predominates ; sometimes
the spark is reduced to two undulations, composing a kind of
S, frequently found in lightning which strikes the ground (188).

282. The crooked form, in particular, which we have already
described in the smaller sparks from a rheostatic machine of
ten condensers, is found near to the negative pole (266). The
shape in which the hook is produced constantly varies. Thus, we
have obtained series of sparks in which the crook is in the opposite
direction to that in fig. 78 but always near to the negative pole
(fig. 79, 286).

(i) When the insulating surface is of pure resin or vulcanite the trace is not nearly so
easily seen.

221 »

283. The formation of this crook appears to us capable ol
being explained by the meeting of the two ponderable matters, in
motion from opposite directions, projected from the points of the
exciter, and by the angle that nearly always results from this
meeting ; for the electric jets are hardly ever produced in a line
with each other ; they start from various points at the ends of the
exciter however fine they may be. Each of these ends, even when
pointed, offers in reality a relatively large surface in proportion to
the extreme fineness of the jet of electric matter which escapes
from it, and the point from which this jet strikes depends upon
the most varied circumstances, either in physical condition or the
chemical state of the surface, as we have had occasion to notice in
the more abundant stream of luminous matter produced by a
high tension current between a platinum electrode and the surface
of a liquid.

As regards the formation of the crook close to the negative pole,
this is accounted for by admitting that the electric movement
starting from the positive pole must be the more rapid of the two,
and that it traverses the greater part of the distance to the other
pole whence there takes place an opposite movement; conse-
quently, the angle or rounded crook resulting from the meeting
must be naturally produced in the vicinity of the negative pole.

284. Arborisations.— These discharges also present arbores-
cence, which appears on dispersing the excess of sulphur by giving

the insulating plate a few light taps.

•i&
Figure &6 represents, in natural size, the arborisation formed

along the course of a discharge fifteen centimetres in length,
produced by the rheostatic machine.

285. These effects account for the vegetable appearance that

222

79-

223

has been sometimes noticed imprinted upon the bodies of persons
struck by lightning, which are only the result of ramifications
in the track of the lightning itself.(I)

This case is easily explained by its analogy with what transpires
in the preceding experiment. At the moment when the discharge
is produced, we see the flowers of sulphur thrown into the air,
especially round the two poles. In the same way, in the case of
the lightning stroke, the dust or any matter placed in track of the
discharge must be dispersed and we can imagine that this matter
raised to a very high temperature, might produce upon the human
body an effect of instantaneous cauterisation in an arborescent

form.

•%
286. Fig. 3$ represents the track produced in flowers of sulphur

by a spark from the rheostatic machine before giving the insulated
plate, on which the powder is spread, the shaking which makes the
arborisations appear.

It may be remarked that the breadth of the track is greater in
the direction of the positive and narrows as it approaches
the negative pole.

Round the positive pole may be perceived some jets correspond-
ing with the branches or rays along which the flowers of sulphur
has been thrown up and scattered in greater quantity.

(i) An instance of this was recently given in the " Lancet:' "A shepherd in

Leicestershire was watching his flocks in the fields when a storm broke out and, naturally,
like most people insist on doing, he sought refuge under a tree. A short time after,
he felt a shock on his left shoulder and, suddenly losing the use of his legs, fell to the ground.
When he was carried home he was still in complete possession of his senses but he com-
plained of pains in the back and the legs. The examination to which he was subjected by
the doctor who was called in, brought to light a rather odd effect of the lightning. From
the left shoulder, for the entire length of his back, there appeared, wonderfully produced,
irregularities upon the skin, and, in a brilliant scarlet tint, a tree-like stem with numerous
branches delicately traced as with the point of a needle. The trunk was about three-
quarters of an inch wide and the general appearance was that of the lower part of a fern with
six or eight branches. The whole was very well reproduced as if printed upon the patient's
back." (Les Mondes, September i2th, 1878).

224

Near to the negative pole, on the contrary, circular tracings are
seen, corresponding with the outline of the arborescent bouquets
formed around this pole, which are, as has been shown in
fig. 80, of quite a different kind from those near the positive pole.

287. Liehtenberg- figures, produced by sparks from
the rheostatic machine. — These sparks, produced on a surface
of pure resin, by means of blowing the powdered sulphur and red

Fig. 80.

lead, produce good figures like those ol Lichtenberg, of a
different kind from the arborizations above mentioned and which,

225 *

fixed upon paper moistened with varnish, constitute valuable
examples for the study of the electric discharge (figs. *8i and 82).
The difference between the effects produced by the sprays and
that produced by sparks is here specially marked.

When the distance between the points of the exciter is too great
for the spark to strike across and only a spray appears, the electric
motion of ponderable matter leaving the negative pole, shown by
the powdered minium adhering to the resin, does not reach the
positive pole. This latter pole shows no trace of red dust in the
midst of the crown of sulphur with diverging rays which surrounds
it (fig. 8 1).

But, if the spark has struck across, this crown is open and the
interior becomes filled with dust of red lead, showing that the
electric motion leaving the negative pole has reached the point
from which the positive electricity starts, fig. 82.

In the case of the spark, the distribution of negative electricity
presents a curious crabiform appearance (fig. 82)(I); in that of the
spray, the electric motion around this same negative pole offers an
aspect, no less strange, of a polypus directing its tentacles towards
the positive pole without being able to reach it (fig. 81).

288. Moreover, when the spark flashes, one may sometimes
know by the traces of sulphur around the negative pole, that the
emission of positive electricity has extended even so far as to
reach that pole. There is then a mixture of the two electricities
at each pole (fig. 83).

*Fig. 81 reduced to three quarters natural size.

(i) This form is not exceptional; it has occurred in a great number of sparks obtained
in the same manner.

226

;Vvv VwVY-T J, - *

^Wilili
^yS^ff!

AfflKlBt.

Fig. 81.

227

Fig. 82.

228

289. This observation explains how, in the circuit of currents
of very high tension which closely resemble a continued series of
discharges of static electricity, a complete decomposition of the
water at each pole may be obtained and, consequently, a mixture
of oxygen and hydrogen (132).

83-

290. It may be also seen that the movement leaving the
positive pole surrounds the negative electric movement as with a
bunch of squibs of a curved trajectory (figs. 81, 82 & 83).

There is often seen, at the same time, an interior flow of positive
electricity round the line of the spark, besides the positive current

229 -

surrounding the exterior, and, between the two, the electric
I negative current which seems as

though it were drawn in by the
positive pole (fig. 84). Negative
electricity, or the ponderable
matter which it carries with it,(1)
moves in an annular space
formed by the electrified mat-
ter coming from the positive
pole.(3>

291. This latter fact explains
the effects of suction or ascension
of water that we have obtained
with electric currents of high
tension (148).

Does it not also explain, as we
have already pointed out, the
+ ascension of water in the cloudy

Fig. 84. substance of water spouts (228) ?

292. Condensation of the Sparks.— Sparks from the
rheostatic machine make a considerable noise ; for they result from
the discharge of a very great number of condensers. But their
intensity may be further increased by charging Leyden jars or

**,

y* /

\ \

/ / +

\
\

1

\v

;' /

\ V

\ \

/ if

\\ \

f \

\

» /

\

/ /

\

/ /i

* '1

*/ /

i j

: |

||

+ 1

• I
1
\ \

I

A \»

- I/

\\

ll

1 i
1

*\ H

+ ll /'•«•

/' /

(1) When we say ponderable matter, we do not mean powdered sulphur and mimium,
but that matter, invisible when cold, drawn from the electrodes by disruptive discharge, the
passage of which on the resin is revealed by the subsequent insufflation of the mixture of
sulphur and red lead.

(2) All these effects may without doubt be observed by means of discharges from Leyden
batteries ; but we here describe them as obtained with the rheostatic machine, not only to
show their identity with those of static electricity, but also because this machine allows us
to reproduce them on a large scale and with greater clearness than by any other means.

230

"batteries. It is only necessary to leave an interval of one or two
centimetres between one of the poles of the machine and one of
the armatures, and letting each charging spark flash across the
space, as with induction coils, in order to prevent any partial
discharge from the jar or battery when the condensers are in
parallel during the rotation of the machine.

Under these conditions jars or batteries may be charged, just
the same as with a powerful electrical machine, and retain their
charge very well.

Five or six sparks suffice for highly charging a large jar. A
battery composed of four bottles is charged in a few seconds ; for,
even by a very quick rotation of the machine a continued series
of powerful sparks is obtained.

293. In using the Lane electrometer, so that the jar discharges
spontaneously and in proportion as the machine rotates, frequent
sparks of 5 centimetres in length strike between the branches of
the exciter with a loud noise.

294. Coloured sparks.— Sometimes, under certain condit-
ions, the sparks are seen to be of a yellow colour much brighter
than the ordinary sparks and analogous to those which have been
obtained by M. Teploff with an electrical machine.'1'

These are the circumstances under which we have remarked
them. The circuit of the rheostatic machine, when in series, is
closed by a Leyden jar of very small surface, formed by a long
tube of thick glass that a single spark from the machine is more
than sufficient to saturate, and the distance between the armatures

(i) Journal de Physique, t. VIII, p. 131, 1879.

231 -

of which (equal to om2o) is such that the discharge cannot take
place.(I) Besides, the exterior armature is formed of two metal
rings about a centimetre apart.

Electric brushes then appear between the first ring of the outer
armature and the unattached rod outside which communicates
with the inner armature. But, at the same time, in the space
between the inner armature and the second ring of the outer
armature, are seen the yellow sparks of which we speak. These
make little noise ; they seem rather to flash between the interior
and exterior of the glass.

295. This kind of spark seems to us to be owing to a partial or
imperfect discharge across the glass, and to the electro chemical
action which results from it.

The discharge is not strong enough to pierce the insulating
substance,'2' and, on the other hand, the tension is such (the
armatures being charged to saturation) that the electric decompo-
sition must be accomplished more or less completely in some way
or other. The result under these conditions is an electrical effect
of quantity condensed on one point, an effect in this case rather
calorific than mechanical, consequently, a chemical decomposition
of an extremely small part of the interposed substance, and, as
this substance is of glass, an incandescence of the sodium which
gives the spark its characteristic yellow colour.

296. The red colour observed by M. Teploff with moist con-
ductors such as wet cords ought to be explained, in our opinion,

(1) If this distance is only from 01114 to 0^15 the spark flashes between the armatures.

(2) The conditions are so arranged that this effect cannot take place, for two sparks are
sometimes sufficient to pierce ajar of small surface formed by a tube of thick glass. It
would be possible nevertheless that a very fine crack or an invisible hole might be produced.

in the same way, by the incandescence of a very small quantity
Of released hydrogen.

In short, glass or other substances would not act here only as
non-conductors but as electrolytes on the surface.

297. We have obtained, in addition, sparks of a red colour, by
charging, by means of the rheostatic machine, condensers with thin
insulated plates of vulcanite. One or two sparks suffice to pierce
them; and the sparks resulting from a partial charge, which is
subsequently obtained, are seen in the form of a little cylindrical
red flame emerging from the two sides of the hole in the conden-
ser and passing about half a centimetre beyond it.

The red colour is, in this case, again owing, in our opinion, to the
hydrogen from the vulcanite, partially decomposed by the passage
of the discharge under these particular conditions. The extreme
tenuity of the hole formed allows of a certain accumulation of
electricity on the armatures and also its egress in a large enough
quantity to produce electro-chemical action.

298. Halo of red light in a vacuum.— Sparks from the

rheostatic machine easily illumine Geissler tubes, as already
mentioned (270), without producing stratifications, unless the
current which works the machine be reduced to a relatively
low tension.

The light surrounds the two electrodes and more completely
fills the tube than the light supplied directly by the 800 secondary
couples. There is no marked difference in its appearance at the
two poles.

But, if air, rarefied by the condensed spark of which we have
just spoken, be made to pass through the tube (293) by placing

233

the latter between the arms of the exciter, in the circuit of which is
placed a Leyden jar, it may be observed that the light produced at
the positive pole is encircled by halo or fringes of bright red.

299. We think this colouring of light in a vacuum may be
explained by the incandescence of hydrogen proceeding from the
small quantity of water vapour of which the glass tube always
contains some traces.

When the spark from the machine (when is series) passes through
the tube, the decomposition does not happen in a perceptible
manner because the quantity of electricity is wanting. But, as
soon as there is condensation, electro-chemical action is produced
and the colouring peculiar to incandescent hydrogen appears.

300. Means of valuing very small quantities of
matter or very short intervals of time by means of

the rheostatic machine. — We have already described (274)
an experiment intended to show that, by means of the rheostatic
machine, a small quantity of dynamic electricity can be completely
exhausted in the form of static effects. A secondary battery was
charged for only fifteen seconds; then it was made to work on the
machine and this feeble charge, when transformed, could illumine
a vacuum tube continuously for more than a quarter of an hour.

According to the number of turns made by the machine and
the number of sparks obtained at each turn of the commutator,
it is found that the action of the primary cells on the second-
ary battery during fifteen seconds, corresponds with the production
of about 10,800 sparks in a vacuum.

It follows that the action of the battery during one second is
represented in this experiment by 720 sparks from the rheostatic

234

machine, or, in other words, the production of a spark corresponds
to a duration in the action of the primary battery of ywth of
a second.

On the other hand, by introducing a voltameter into the circuit
of the primary battery, we have discovered that the charge taken
by the battery and yielded by the rheostatic machine in the form
of those 10,800 sparks, corresponds with a consumption of 18
milligrammes of zinc in the primary battery.

It follows that the solution, or deposit, of one milligram of
metal may be thus proved by the production of about 600 sparks
from the rheostatic machine, or, in other words, that the production
of one spark corresponds with the consumption of about Trerrth of a
milligram of metal in the primary battery.

Then, by taking the electric sparks as units, there would be a
means of measuring either very small quantities of metallic
matter dissolved, or deposited, by electro-chemical processes in
very short intervals of time.

It is, besides, easy to determine the number of sparks produced
in air or in vacuum by the rheostatic machine ; because we know
the exact number of sparks made by each turn of the machine.
The machine may be also made to turn as slowly as one wishes
and to expend in any length of time, in a static form, the effect
produced in an extremely short time by dynamic electricity.

301. We may likewise infer from these experiments that the
feeble static tension effects shown directly by the poles of a battery
composed of a great number of elements must not be considered,
as was at first thought,(l) as independent of the electro-chemical

(i) See Gassiot, Philosophical Transactions, 1844.

235

action produced in the battery ; on the contrary, they correspond
with real electro-chemical expenditure, doubtless very small, still
not absolutely nil, even when the circuit seems open and it is
question of but a single spark given with the electroscopic con-
denser, or of an act of simple attraction or repulsion. (1)

302. Effects of quantity from the rheostatie machine.

The rheostatie machine can also give static effects of quantity
which materially differ from those of intensity, by combining all the
condensers in parallel and fitting it with another small special
commutator intended to collect the discharges without confusing
them with the effects of the secondary battery.

This commutator is formed by a little cylinder in vulcanite on
which are four copper bands as at m n and op, placed in pairs
opposite each other, against which the six springs B C E B'C'E',
rub, also in pairs, opposed to each other (fig. 85, 86).

The pair of springs B B' communicate with the secondary
battery ; the pair C C' with the two charging poles of the rheostatie
machine above described (260, fig. 74), previously turned into a
position so that the condensers shall be combined in parallel.
The pair of springs E E' communicate with an exciter or any
other apparatus through which the discharges are required to pass.

(i) It is difficult to conceive at first sight the production of electro-chemical effect in
a battery the circuit of which is not completely closed. But it must be considered that
when the two poles of a battery are put in communication with the armatures of an electro-
scopic condenser, it exercises across the insulating substance a peculiar action equivalent to
an imperfect passage, which necessarily involves a chemical expenditure in the battery.
We have seen, in fact, that if the battery is at a high tension, the charging sparks are
perceived at the moment when the battery is put in connection with a system of condensers
and it can be even proved that there is a decomposition of the water in a voltameter placed
in the same circuit as the condensers (278). These effects show that there is a passage (to
a certain degree) of the current from the battery ; accordingly, it must produce in the interior
a corresponding chemical action. It is the same in a lesser degree if one of the poles of
the battery is put in connection with an electroscopic condenser, the other pole being in
connection with the ground, or even if a substance touching the ground is only brought near
to one of the poles, for there is always an interval between the poles of the battery filled
with the surrounding air which plays the part of the dielectric medium in a condenser.

236

In the position of the commutator represented by fig. 85 the
two poles of the battery B B' are connected by tongues, as at ;;/ n,
with the two poles of the rheostatic machine C C', and the con-
densers then become charged simultaneously in parallel.

In the position of the commutator represented by fig. 86 the
poles of the battery are outside any circuit and the poles of the
rheostatic machine, charged in parallel, communicate by the two
opposite tongues, as at o p, with the springs E E' in the circuit
of discharge.

B',

Fig. 86.

This commutator may be rapidly rotated in order to produce a
nearly continuous series of static discharges in parallel for quantity.

Thus, whilst in the rheostatic machine before described all the
condensers are charged in parallel and discharged in series, in this
case the condensers are charged and also discharged in parallel,
immediately after their charge, by the secondary battery without
the current from that battery interfering in any way with the
circuit of discharge from the condensers. (I)

(i) Among the experiments or arrangements of apparatus with which the rheostatic
machine when arranged for quantity may be analogous we mention those of M. Werner
Siemens who had obtained permanent galvanometrical deviations with condensers composed
of different insulating substances successively charged with the use of a vibrating balance
by some Daniell elements and had determined in this manner the inductive power of these
substances. (See G. Wiedemann, Galvanismus, 2nd edition, vol. I, p. 199. — E. Mascart,
Traite d'electricite statique, vol. II, p. 400).

237

303. The commutator which we have just described, instead
of being arranged with a special driving gear near to the rheostatic
machine, may be fitted to the machine itself, as at a b' (fig. 87)
so that it may be put in movement by the rotation of the machine
independently of the commutator for quantity and tension, a b
being kept at rest in a position which unites in parallel all the
condensers by pressing the button K.

Fig. 87.

When, on the contrary, effects of tension are wanted, the button
K must be loosened and the next one pressed which unites the
axis qf the two cylinders a b a b' , and the wires from the battery
must be connected with the terminals of the long commutator.
The two cylinders turn together ; but the shorter does not fulfil
any function, the first unites the condensers successively in
parallel and in series.

304. If one sought to obtain only effects of quantity, the
cylinder a b should be taken off and the machine reduced to the
small commutator a b' under which should be placed a vertical
or horizontal battery of condensers connected in parallel.

238

305. Static sparks Of quantity. — By putting the rheostatic
machine, arranged as we have just described, in communication
with a secondary battery of from 400 to 800 couples(l), and by
rapidly rotating the little commutator, a continued series of noisy
sparks is obtained, but very short (rV to i\ of a millimetre),
presenting the appearance of a very brilliant point surrounded
with a halo and throwing forth rays of particles from the electrodes.

This kind of spark, though in some respects similar to that
of induction, has a character peculiar to itself and produces
different effects.

(i) In order to facilitate the means of trying experiments that can be made with the
rheostatic machine for quantity or tension, we here indicate a simple means of charging
a battery of secondary couples designed to work the machine without the necessity of using
special batteries with commutators composed of numerous springs for charging and dis-
charging.

These couples may be reduced as we have already explained (yol^-f; note 273), to small
thin plates of lead a few millimetres only in breadth. They should be bent in the shape of
forks, or tuning forks, keeping a certain length in the upper part to be able to take hold of
them easily.

Fig. 88.

A hundred of these little couples are charged at once by placing them in a long narrow
tray made of gutta percha, in which there are two compartments about 50 centimetres in
length. The couples are placed outside on the separating partition as shown in fig. 88
where, for the purpose of explaining more clearly, only a few are represented. Two long
thin plates of lead act as corresponding electrodes which are connected with four Bunsen or
six Daniell elements. The electro motive force necessary to charge this system ought to be,
in fact, double that sufficient to charge a single secondary cell.

The small plates are charged, if care be taken to/orm them as we have pointed out (53), by
changing the direction of the current several times and they end by preserving a large quantity
of the charge received. Directly they are charged they should be plunged into closed glass
tubes each very close to the other, and the little. secondary couples thus formed suffice for
making numerous experiments with the rheostatic machine.

239

The halo is much more developed, especially at the upper part
of the luminous point, and is visible without the necessity of
insufflation. It forms a crown from 8 to 10 millimetres in diameter.

306. In spite of their static origin and the sharp noise they
make, these sparks are of a less tension than those from the
secondary battery itself.

Their length is, in fact, less than that of a spark produced
directly through the air by a secondary battery of 800 couples (140).
Moreover, they do not illumine vacuum tubes, as does the current
coming straight from the battery. To obtain light in vacuum
with these sparks, the distance between the electrodes must be
reduced to one or two millimetres.

307. This inferiority in the tension of the spark of quantity
from the rheostatic machine, in comparison with that of the source
of electricity which charges it, is explained by comparing the
charge of a condenser with that of a voltameter or secondary
couple.

The opposing electro motive force yielded by an accumulator
of electrical work, whether it be a condenser or secondary cell,
could not be greater than that of the electric source itself.

If, in the case of a secondary couple, the principal effect is
chemical action on the electrodes and, consequently, the electrolysis
of the interposing liquid, there is, in the case of the condenser, a
corresponding physical action exercised on the insulating medium
which separates them.

From this cause, in both cases, there is a loss of E.M.F. in the
current during the charge and an opposing E.M.F. given by the
accumulator necessarily inferior to that of the charging current.

240

308. The greater noise produced by the spark of quantity
from the condensers of the rheostatic machine, compared with
that made by a spark direct from the secondary battery may be
explained in the following manner :

The work effected by a current of dynamic electricity of high
tension on a condenser may be considered like putting two
opposed surfaces of insulating matter in vibration, reaching
to a certain depth depending on the tension of the source or the
nature of the substance. This vibration continues a certain time
after the charge, like that which would occur by a purely mechan-
ical action on any sonorous substance. From the moment of the
discharge, the movement being suddenly counteracted by the
speedy return of the matter to its natural state, there results a
peculiar noise in the whole of the circuit of discharge, in the
spark and even in the insulating matter. Now, one can under-
stand that this sudden decomposition of the compact molecules
•of a solid substance may be accompanied by a sharp noise, quite
different from that resulting from opening or closing the
circuit of a high tension battery in which there is no solid
insulating substance, but a series of liquid conductors which form
the principle part of it.

Thus, the noise of the static spark of quantity, of which we
speak, compared with that of the spark direct from the secondary
battery, in spite of a lower tension, appears to us to be the result
of the particular nature of the discharge, or, more exactly, of the
nature of the insulating matter to which the electric vibration has
been communicated during the charge.

309. Calorific effects. — The calorific power of these sparks
of quantity from the rheostatic machine is naturally greater than

241 >

that of sparks of tension from the same apparatus. Platinum or
steel wires, from ro to 20 centimetres in length and from yV to *V
of a millimetre in diameter, may be reddened or melted, whilst the
longest sparks of tension would pass along them too easily to
produce any perceptible heat.

310. Mechanical effects. — The mechanical effects produced
are very strong. If these static sparks of quantity be passed
through a voltameter filled with a solution of salt, of which the
negative pole is a Wollaston electrode, and where the long sparks of
tension would run through silently, the passage of the former is
accompanied by a very loud sound similar to a small explosion ;
the mechanical effect produced is so strong that even the vessel
of the voltameter is displaced and pushed forward on its stand ;
the glass begins to vibrate, and, if the commutator be quickly
turned, loud rolling or ringing noise is the result..

311. By disposing the connections so that the secondary
battery acts at the same time on the voltameter, by the inter-
mediary of imperfect contact, continued breaks are produced
spontaneously, and the ringing sound becomes automatic. Any
rhythm given to these interruptions is repeated with great intensity
in the voltameter and it might be possible, perhaps to make use
of this in the telephone.

312. Rheostatic hydraulic ram.— The greater part of the
phenomena that we have observed in employing currents of high
tension, are shown with greater facility and less tendency to be
transformed into calorific effects, by the aid of these continuous
discharges of semi-dynamic semi-static origin. The experiment we
have described by the name of "voltaic pump" (148) is here repro-
duced very clearly by an altogether mechanical action of electric

242

force. Instead of rising uninterruptedly, as with a continuous
current, the water ascends by jerks, more closely together
as the sparks follow each other more quickly, and the apparatus
then becomes a true rheostatic hydraulic ram.

By a continual interruption of the current, spontaneously pro-
duced as in the preceding case, the effect also becomes automatic.

Fig. 89.

313. Vibration knots formed in a metallic wire by
the current of quantity from the rheostatic machine.—

The passage of the current of quantity from the rheostatic
machine through very fine platinum wires (of TO of a millimetre
in diameter) is accompanied by remarkable machanical effects.

As soon as the machine is turned, sharp angles appear through-
out the length of the wire (about om4o), at semi-regular distances,
forming a series of accolades or vibration knots. The wire,
which was half stretched, rises and changes from the form a b
to the form a'b' (fig. 89).

243 „

These angles appear to be nearly equal distances apart, but
sometimes two or three consecutive ones are seen with the angle
pointing in the same direction.

If the machine be again turned, after having placed the nippers in
which the wire is held nearer together so that it does not stretch
to the point of breaking, new bends appear round the angles
already formed and the wire becomes like a'b" . If it is short-
ened to omio in length it becomes white hot, presenting
numerous angles and sinuosities so sharp (d"b"') that it has the
appearance of a continuous electric spark.

In the latter case, it is found shortened after the experiment
to the extent of 5 or 6 millimetres in the length of 10
centimetres.(l)

314. There is occasion to remark, that, if the platinum
wire is new and annealed by the wire drawer, and if it has
not been previously reddened by any kind of calorific source or
current, it forms these knots of vibration much less easily. This
shows that the current must contend with molecular cohesion to
produce this phenomenon.

(i) These phenomena may be compared with those which have been observed by Nairne
and by M. Edmond Becquerel with discharges from Leyden batteries, and with those ob-
served when a long fine wire is made red by a battery composed of a great number of
elements. But here they are more marked and present other characteristics by reason of
the different nature of the electric source employed which possesses at the same time the
dynamic and static state by the quantity and the tension of the electricity in play.

Nairne had observed that metallic wires submitted to discharges of static electricity
underwent a diminution in their length.

M. Edmond Becquerel found that this diminution was inversely proportioned to
the cube of the diameter of the wire, and has further observed that the wires become
undulated by the action of discharges given by two Leyden batteries of nine jars. The
undulations increase in height, according as the discharges follow each other, without ever
disappearing to make room for others. (See Traite d'electricite in 3 vol. by MM. Becquerel
vol. I, p. 309).

244

315. Variations of the distance of the knots accord-
ing- tO tension Of the Current. —The distances at which
these knots or sharp angles appear so clearly from the first instant
of the current passing, represented at a b' (fig. 89), do not
depend on the rate of rotation of the rheostatic machine. If the
rate is great it only makes the transformation of the current from
the secondary battery less complete.

But these intervals between the knots appear to depend on the
tension of the current. Thus, by diminishing the number of second-
ary couples working on the rheostatic machine to half; in reducing
them from 800 (the number used in the preceding experiments)
to 400, so as to have a very decided difference in the tension of
the current of discharge given forth from the machine, we have
found that the distances between the knots, which were in the first
case from i to 2 centimetres,'1' vary in the second case from 2 to
3 centimetres.

The amplitude of this kind of longitudinal vibration, produced
by this peculiar electric current, seems to increase according as
the tension of the current is diminished.

316. Noise in the Wire. — While this phenomenon is
happening, there is heard around the wire a noise or cracking
sound similar to that which would be produced by a spark,
although in this case the wire has no break in it.

317. This noise in the wire without the intervention of any
electro magnetic action is an important point for consideration.

(i) These knots are not formed at perfectly equal distances as we have before remarked
(313). They generally succeed each other in couples with a shorter interval between as they
continue. Thus, after two intervals of nearly two centimetres, there appears an interval of
not more than one centimetre and so on. The investigation of the law which governs these
divisions of the wire would merit separate study.

245

It can only be owing to the molecular disturbances resulting from
the passage of the current peculiar to the machine, which causes
these sudden contractions and distortions in the body of the
substances through which it passes.

This phenomenon shows that a corresponding mechanical effect
must be produced inside the insulating matter of the condensers,
which are here the source of the current which passes along the
wire, just as the calorific or chemic effects observed in the exterior
current of a battery are but the reproduction of those which occur
in the interior of the battery itself. That would then be the cause
of the noise or sound made by a condenser at the time of its
charge or discharge.

318. While the wire undergoes these vibrations of a longi-
tudinal appearance, the effect of which is permanent and remains
visible, it is subjected also to transverse ones of very great size
which strongly disturb it.

319. The longitudinal vibrations are a mechanical effect of
the current which must not be attributed to its disconnected
nature ; for, if the current direct from the secondary battery
(rendered disconnected by the same commutator acting as simple
interrupter) is passed in the same wire no permanent change can
be ascertained in the form of the wire.(I)

The transverse vibrations, on the contrary, result from the
calorific effect produced by the current being alternately continu-
ous and interrupted ; for, if the current from the battery, in an
intermittent manner as above, be made to pass along the wire these

(i) Care must be taken in this case, by interposing a column of water, to greatly diminish
the quantity of the current from the battery without perceptibly weakening its tension, in
order to prevent it reddening or melting the wire.

246

vibrations are seen, although less strongly than with the current
from the rheostatic machine. In this latter case, calorific effect
is produced very abruptly and stops in like manner, by
successive discharges from the condensers. The consequence is
that the wire falls while heating, rises while cooling, and violently
trembles under the influence of the current.

320. Fragility Of the Wire.— The wire becomes very
brittle after the current has passed through it. If the experiment
lasts more than two minutes it breaks spontaneously.

This tendency of a wire to become brittle under the influence
of an electric current had already been remarked by Peltier and
other observers. But, with ordinary currents of dynamic electricity,
it was so slight that it was not altogether admitted.'1' Here it is
evident by reason of the peculiar nature of the current employed.

321. Results relative to lightning conductors.— If the

discharges from this machine, by passing through a fine metallic
wire, can produce such a change in the molecular structure
as to cause it to break spontaneously after some instants, the
passage of the currents from the lightning, which combines in
a much greater degree electrk quantity and tension, must produce
similar effects on much thicker conductors such as the rods or
iron cords of lightning conductors.

These conductors may then become very brittle and present
invisible ruptures, not only after the direct fall of the lightning
that they have been able to carry, but also when they have
served for a long time for the silent stream of large quantities

(i) See "Resume de 1'histoire de 1'electricite et du Magnetisme" by MM. Becquerel,
p. 237.

247 *

of atmospheric electricity. They may even have received a
certain number of discharges without there being any appreciable
interruption discovered by the aid of electric instruments and yet
be in such a state of molecular fragility that another powerful
discharge may achieve the rupture of the conductor, as in the
experiments above described.

Thus, accidents may be accounted for which happen with
lightning conductors apparently quite sound. (I)

322. Conclusions drawn from the preceding1 phen-
omena relative to the mode of propagation of

electricity. — The phenomena we have just described (313 to
320) are of a nature to throw some light on the mode of
propagation of electricity. The molecular vibrations revealed
by knots formed in a metallic wire, by the curious noise and by a
notable change in its cohesion under the influence of the passage
of the dynamo-static current which we have just studied, must be
produced in a lesser degree in conducting substances traversed by
electric currents of very low tension. This vibration may be
too feeble to be perceptible but it is not the less real.

We are then able to conclude that the electric movement must
diffuse itself in substances after the manner of a purely
mechanical motion, by a series of very rapid vibrations of the
more or less elastic matter through which it passes.

These facts may be compared with those we have observed
when a current of high tension passes above the surface of water,

( i) It may be necessary to completely renew a lightning conductor, even though it seem
to possess sufficient conductibility, if it be often exposed to frequent violent storms so as to
become brittle in consequence of the great quantity of electricity which passes through it.
M. Callaud has often had occasion to observe that the cables of lightning conductors
become very brittle, and has attributed this fact to the passage of electricity. (Traite des
paratonnerres, par A. Callaud, p. 91).

248

and this, being in vibration, presents to view a number of remark-
able luminous forms which recall to mind those made by vibrating
plates in acoustic experiments (138).

In these phenomena are found novel analogies between electric
and vibrating sonorous motion which is itself a mechanical motion
of ponderable matter.'1'

323. On the quality of reversal possessed by the

rheostatie machine. — If, instead of passing a current of
dynamic electricity into the rheostatie machine to obtain static
electric effects, the machine be put in connection with a direct
source of static electricity, such as an electric machine or
another rheostatie machine at work, tokens of reverse transforma-
tion are obtained, that is, traces of dynamic electricity.

In this case, the tension poles of the rheostatie machine are put
in connection with the static electric source and the charging
poles, which are joined to the battery, communicate with a
galvanometer.

If the commutator of the rheostatie machine be at rest in
such a position that all the condensers are in series, the electric
machine charges them in tension, or in series, even though they
may be a large number, because of the thinness of the mica plates.
When the commutator is revolved, the needle of the galvanometer
marks with an abrupt movement the direction of the dynamic
current produced by the discharge of all the condensers combined

(i) We had already been led to these conclusions after observing effects produced by
electric currents of high tension and we had also considered electric discharge as a mechani-
cal movement ; but more particularly as an extremely rapid transporting movement of a
very small quantity of animated ponderable matter. (Comptes rendus, t. LXXXVIII,
p. 442. Extrait d'un pli depose a 1'Academie des sciences, June nth, 1877 and opened
March 3rd, 1879).

249

in quantity. This collection of condensers, with very thin
insulating plates possessing a large surface, plays the part of a
voltaic couple which would nevertheless possess a powerful resist-
ance, because of the nature of the medium which separates
the electrodes.

If an electric and a rheostatic machine are set going at the
same time, the electric machine will not charge the condensers
fast enough for a perceptible deviation of the needle to be
observed ; but the introduction into the circuit of a telephone, as
galvanometer, reveals by a rustling noise the passage of a broken
current of small intensity.

In the case of two rheostatic machines being used, one as static
electric source, the other as receiver, to obtain an inverse result
the effects are more marked.

The rheostatic machine may then be considered as reversible,
like most of the machines purposed for the transformation of
force. But the generators of static electricity furnish so weak a
quantity that, even supposing it to be completely transformed into
dynamic electricity, the current obtained would have too little
intensity to be advantageously utilised. We restrict ourselves to
mentioning these results as giving another example of the bonds
which exist between the different modes of manifesting electric
motion and to show the possibility of transforming one into the
other by the most varied means. (1)

(i) Paragraphs 275 to 323, inclusive, are taken from a note published on the i4th July,
1879, (Comptes rendus, t. LXXXIX, p. 76 a 80), and from a pamphlet presented to the
Academic des sciences, the 6th Oct., 1879. (Comptes rendus, t. LXXXIX, p. 605)

$IXTB PART.

Analogy between electric phenomena
and effects produced by mechanical
actions. Results relating to the
nature of electricity.

" Convertenda plane est opera ad inquirendas et notandas
rerum similitudines et analoga, tarn in integralibus quam partibus :
illae enim sunt, quae naturam uniunt, et constituere scientias
incipiunt." (Bacon, Novum Organum^ lib. II, § 27).

324. Since the first observations we made, by subjecting sub-
stances endowed with great molecular mobility, such as liquids,"1
to the action of electric currents of high tension, we have been
struck with the analogy which the phenomena produced presented
to those which result from the action of mechanical force,
correctly speaking, on the same substances, particularly when this
force is represented by an extreme rapidity of motion communicated
to a small mass of matter. (2)

(1) Comptes rendus, May $th and July 26th, 1875.

(2) We have already pointed out a certain number of these analogies in the course of
our researches ; but we now think it useful to collect them and draw from them some
conclusions.

251

The relations which exist between the two classes of phenomena
have appeared to us more visible in these experiments than in
those by static electricity because we put in play a greater quantity
of electrified matter; and more evident also than with ordinary
electric currents because we employed a much higher tension.

But when once these analogies are discovered, they are easily
recognized in a more or less marked degree in nearly all pheno"
mena produced by static or dynamic electricity.

325. If, among the phenomena that we have described some
are chosen in which these analogies appear most striking, in the
first place may be cited the phenomenon of the luminous waves,
produced in the heart of the liquid around the extremity of an
electrode pressing against the side of a voltameter, through
which an electric current of high tension passes out (157).

The violent movement imparted to the liquid, being stopped
by the sides of the voltameter, raises the glass to a temperature
high enough for the luminous circles to be formed/1' and, when the
current is of sufficient quantity and tension, luminous waves are
formed in the midst of the liquid itself.

There is then found in these phenomena a representation of

the waves formed on the surface of a liquid by the fall of

a mass of solid matter, modified only by the production of
calorific and luminous effects.

The varied luminous shapes, produced by a current of yet
greater tension striking the surface of a liquid (138), are quite

(i) In this experiment the waves produced remain imprinted on the glass in the form of
concentric rings.

252

analogous, in form, to those which are caused by the fall of
liquid drops on the surface of a liquid. If the figures we have
observed be compared with those which have been obtained
by Mr. Worthington and others'1' there will be found among them
some which are nearly identical. We have also mentioned the
analogy of these shapes with those which result from the vibration
of sonorous plates (II, 322).

We recall to mind that the stratifications of electric light in
vacuum, observed by MM. Abria, Grove, &c., which present
very decided curves in receivers of sufficient diameter, have been
already compared by Gassiot, de la Rive, &c., with waves produced
by mechanically shaking a liquid.

326. The phenomenon of the sheaf of finely divided water,
produced by an electric current of high tension (143), has its
analogue in the mechanical sub-division of a liquid by the action
of a jet of condensed air only acting at once on a very small
portion of its surface (Appareils pulverisateurs ).

327. The phenomena of suction, produced by the flow of an
electric current of high .tension (pompe voltaique, belier rheo-
statique) are analogous to those which result from the passage, in
a narrow tube, of a liquid or jet of steam impelled at a great
speed. (Tube de Venturi, Giffard's injector.(a))

(1) Proceedings of the Royal Society, XXV, p. 261, (1876). — La Nature, 5th year. — and
semestre, p. 236, September, 1877.

(2) There is also a close analogy between these phenomena and the effects of suction
recently discovered by M. D. Colladon in waterfalls: — "There may be perceived little
sheaves, composed of millions of liquid pearls impelled at a rapidity of motion absolutely
incredible, in a contrary direction to the water of the cascade, and quickly ascending to-
wards the summit." (Comptes rendus, vol. LXXXIX, p. 286. — Les Mondes, September
25th, 1879, P- i47)-

253

328. The mechanical, calorific and chemic action produced
at the same time by an electric current of a certain tension on the
surface of glass, which has led us to the " engraving on glass by
electricity" (161), may be compared with the action exercised on
this substance by an exceedingly fine jet of sand shot forth under
strong pressure, which has been employed for several years in
America for engraving on glass.

329. The bubbles, blisters or luminous beads formed along a
column of matter through which passes a strong electric current
at the time of fusing a wire (§ 83, fig. 23), or the incandescence
of a narrow stream of rarefied air by a powerful discharge of
atmospheric electricity (188, Eclairs en chapelet), offer a close
resemblance to the phenomena which accompany the flow of
a liquid through a narrow opening under considerable pressure.

330. The vibration knots formed in a wire by the electric
current, either dynamic or static, that we have previously studied
(313), show, as we have already explained, the analogy which exists
between the electric current and the vibratory sonorous motion
caused by a purely mechanical action.

331. The experiment described (271), which consists in
exhausting a tube of rarefied air, so as to promote in it a
luminous stream of electricity, by increasing the tension of the
current of the secondary battery by the momentary addition of
that from the rheostatic machine, is analogous to the effect well
known in the laws of fluids, which consists in exhausting a syphon
and promoting the flow of the liquid by suction.

This analogy may be rendered still more striking. Simply
bring a stick of electrified resin or ebonite near to the tube and

254

suddenly take it away so that the light immediately appears in the
tube. There is thus produced, at the extremity of one of the
electrodes, a kind of suction which, added to the already high
tension of the current causes the appearance of a luminous stream.

When this phenomenon is more closely studied, the effect is
found to be, principally, at the positive electrode. With a
substance, on the contrary, charged with positive electricity, the
illumination is caused by approaching the negative electrode.
In an electrified substance possessing a rather large surface
this effect is very strong ; and we have thus been able to cause
the continuous illumination of a Giesler tube by a secondary
battery of 700 couples, by bringing the metallic plate of an
electrophorus to the distance iIU5o from it.(I)

332. If the best known effects of static or dynamic electricity
be considered from this point of view, numerous analogies are
found with effects produced by mechanical force, especially when
they are compared, as we have done, with mechanical actions in
which velocity plays a more important part than the mass of
matter in motion.

Thus, there is a great similarity between perforations produced
by electricity and those produced by projectiles impelled at a high
velocity; also between calorific effects obtained by electricity and
those resulting from (so called) mechanical impact, and between
gyratory reaction motions produced by a flow of electricity (electric
whirls, &c.,) and those worked by water, steam or condensed
gas (turbines, &c.)

(i) These phenomena may be compared with those observed by William Spottiswoode and
Fletcher Moulton in their study of the sensitive state of electric discharges through rarefied
gases where numerous examples are found of induction effects on electric light in vacuum.
(Philosophical Transactions, 1871, first part, page 165.)

255 *

333. Effects are obtained by mechanical means, from the
division of matter reduced to its first principles, similar to effects
of the same kind caused by electricity, in employing, as force,
substances impelled at a high velocity.

A jet of steam projected under strong pressure against the slag
of blast furnaces, divides it into numberless threads, forming it
into a kind of mineral wool. In the same way, matter impelled by
electric movement sub-divides, to an infinite extent, all other
matter it finds in its way.

334. Cutting tempered steel with an iron disc has latterly
been achieved ;(I) now, during this operation "it throws out a
continuous jet of sparks and particles of steel apparently at white
heat ; nevertheless, the hand may be passed through this jet with
impunity and a sheet of paper interposed is neither burnt nor
blackened. These particles appear to be in a spheroidal state ;
when cold they take the form of an elongated cone resembling
stalagmites ; the steel has been really melted."

In these mechanical effects new analogies are found with the
mechanical, calorific and even physiological phenomena of elec-
tricity. The ordinary electric spark, in spite of its high temperature,
does not burn, by reason of the small quantity of ponderable
matter in play ; the steel also in this case^though melted, does not
burn because of its extreme state of sub-division and, in conse-
quence, its growing cold so rapidly.

335. The phenomena of attraction and repulsion which seem
so characteristic of electricity can be imitated with the aid of a

(i) This result has been obtained by the aid of a machine constructed by M. Jacob
Reese. See Bulletin de 1' Association scientifique de France, November 6th, 1876, p. 77.)

256

strongly compressed jet of air escaping through an extremely
narrow opening. Balls of different substances, even metal, may
be held in equilibrium, attracted or repelled, by a jet of air at
high pressure, according to their distance from the opening,
density, &c.(l)

The recent works of M. Bjerknes have shown the possibility of
also obtaining, by other purely mechanical means, attractions and
repulsions similar to those caused by electricity. <2)

MM. Dvorak and Mayer have observed, on the other hand,
peculiar phenomena of repulsion at the approach of bodies in
vibration.'3'

336. Conclusions relating1 to the nature of electric-
ity.— The analogies which we have just enumerated permit us,
we think, to consider electricity as a purely mechanical motion of
ponderable matter.

This movement consists in the extremely rapid flow, or
transport, of a very small quantity of matter, in regard to the
electric spark, the voltaic arc or electrical discharge in general.

337. Electric motion may occasion gyratory movements the
same as mechanical motion — correctly speaking — by an effect
from reaction due to the flowing of matter, however small the
quantity may be, which escapes from electrified substances. (4)

(1) These experiments were repeated at the International Exhibition in 1878 in the
American section by the aid of reservoirs of condensed air from the Westinghouse brakes.

(2) Comptes rendus, LXXXVIII, p. 165 and 280 ; LXXXIX, p. 134 ; 1879.

(3) Philosophical Magazine, 5th series, VI. p. 225, and Journal de Physique, VIII,
p. 25 ; 1879.

(4) This matter is not electric matter as it was formerly believed to be, but electrified
matter borrowed both from the substance itself from which it detaches itself and from the
Centre through which it passes,

257-

338. Electrical motion may become vibratory like mechanical
motion when the ponderable matter resulting from the discharge,
coming in contact with a substance of a peculiar elasticity, permits
it to transfer the shock received through its entire mass.

This peculiar elasticity constitutes electric conductibility.
There is not then any transport of ponderable matter throughout
the length of the conducting substance, but diffusion by vibrations
similar to those of the sonorous motion or the movement trans-
ferred to a series of elastic balls. The phenomenon of the jet of

ponderable matter may be also produced at the extremity of the

«»f cu^^te iw tV>c >*^t«Vta)

conductor when there is a break^in the middle of the wire.

339. This transformation into vibratory motion may take place,
to a certain degree, in the electric discharge itself, through an
imperfectly conducting medium such as ordinary or rarefied air.
There is then both transport and vibratory motion.'1'

340. The very rapid movement of ponderable matter, which
constitutes electric discharge, produces as we have before said (3 2 7),
like the rapid motion of a fluid, a suction or inverse motion in
the particles of matter which receive the electric shock, or of that
which forms the centre of the matter traversed by the discharge.

From that cause a double movement occurs in two different
directions ; consequently, a double transport of ponderable matter.
To this double movement are due the effects produced in electric
discharge which are, by general consent, called positive and negative
electricity. Instead of these expressions, which seem to infer two
sorts of electricity, the terms "direct electric motion" and "inverse
electric motion" may be substituted.

(i) It is this double effect which often gives to electric phenomena such complicated
appearances.

258

341. As to the phenomena produced by electricity called static,
we consider them as due to the vibratory state of the molecules
at the surface of electrified substances, accompanied by a more or
less abundant emission of material particles detached from this
surface, according to the conditions in which the electrified sub-
stances are placed in reference to the surrounding medium. (I)

The phenomenon of the aigrette is a characteristic manifesta-
tion of this emission of ponderable matter. The aigrette is
always produced in a greater or less degree on different points of
a strongly electrified substance; the least wrinkle in the surface will
occasion it. This phenomenon then reveals the state of continual
discharge in which a substance may be when charged with
static electricity.

It may be also said that this emission becomes more evident
the nearer the electrified substance chances to be to another
substance, not electrified, which serves in some degree as target for
the projectiles formed by the molecules from the electrified
substance.'21

342. To sum up in a few words the views herein stated ; we
think that electricity may be considered as a movement of ponder-
able matter — a movement of transport given to a very small mass
of matter, impelled to an extreme velocity, when there is question

(1) The earlier electricians, principally Boyle, Hankshee, &c., had already allowed that
material effluvium escapes from electrified substances. This idea appears to us to be still
correct at the present time by adding to it the vibratory molecular motion of the surface of
these substances.

(2) We have several times had occasion to point out, in our examination of currents of
high tension, calorific and luminous effects resulting from these molecular shocks. The
experiments made by Mr. Crookes also present numerous and brilliant examples of this kind.

259

of electric discharge — and a very rapid vibration of the molecules
of matter when touching its transmission to a distance in a
dynamic form, or its manifestation in a static form on the surface
of substances. (Oct. i6th, 1879.)

INDEX.

FIRST PART.

THE ACCUMULATION AND TRANSFORMATION OF THE ENERGY OF
THE VOLTAIC BATTERY BY MEANS OF SECONDARY CURRENTS.

CHAPTER I.

The study of secondary currents.

Page.

Historical. — Secondary currents. — Voltaic polarisation ... ... 1

Study of secondary currents produced by various voltameters. — Apparatus

used ... ... ... ... ... ... ... 5

General results of the above ... ... ... ... ... 7

Voltameter with copper electrodes and water acidulated by sulphuric acid 8

Voltameter with silver electrodes ... ... ... ... ... 12

Voltameter with tin electrodes ... ... ... ... ... 14

Voltameter with lead electrodes ... ... ... ... ... 14

Voltameter with aluminium electrodes ... ... ... ... 18

Voltameter of iron and zinc ... ... ... ... ... 19

Voltameter with gold electrodes ... ... ... ... ... 20

Voltameter of platinum electrodes ... ... ... ... 21

Voltameters with a saturated solution of bichromate of potash ... 24

Conclusions ... 26

262

CHAPTER II.

Storage of the energy of the voltaic battery by means of
secondary cells with lead plates.

Deductions drawn from the preceding researches ... ... ... 29

Secondary cell with coiled lead plates ... ... ... ... 30

Secondary battery of large surface for quantity ... ... ... 31

Secondary cells with parallel lead plates ... ... ... ... 31

Another form of secondary cell with lead plates ... ... ... 34

Chemical actions produced in secondary cells with lead plates... ... 38

Formation in electro-chemical preparation of secondary cell with lead

plates ... ... ... ... ... ... ... 41

Absorbtion of the gases during the charging of secondary cells ... 46

Maintenance of secondary cells ... ... ... ... ... 48

Effects produced by secondary cells with lead plates... ... ... 48

Calorific effects... ... ... ... ... ... ... 49

Magnetic effects ... ... ... ... ... ... 50

Formation of o?,one in secondary cells and voltameters with lead plates 51

Shades or tints of oxide produced at the positive pole during the discharge

of secondary cells ... ... ... ... ... ... 53

Duration of the discharge in secondary cells ... ... ... 54

Constancy in the secondary current during the discharge ... ... 55

Preservation of the charge taken by secondary cells ... ... ... 57

Residual charge afforded by secondary cells ... .. ... 59

Increase of the intensity of a secondary cell with rest after being charged 60

The electro-motive force of lead plate secondary cells ... ... 61

Resistance of lead plate secondary cells ... ... ... ... 63

Necessary E.M.F. of the primary current for charging the secondary cells 64

Limit of charge that secondary cells may take ... ... ... 65

Secondary cell charged by a thermopile ... ... ... ... 66

Secondary cell charged and discharged by means of the Gramme machine 66

Various analogies presented by secondary cells ... ... ... 68

Return obtained from secondary cells ... t . . . ... ... 70

263

CHAPTER III.

Transformation of the energy of the voltaic battery by means of
lead plate secondary batteries.

Page.
Historical. — Various means used in order to increase the E.M.F. of the

battery ... ... ... ... ... ... ... 73

Secondary battery for tension effects formed of parallel lead plates ... 75

Secondary battery for currents of high tension formed of couples com-
posed of coiled lead plates ... ... ... ... ... 77

Effects produced by secondary batteries composed of lead plates ... 82

Large secondary battery of extended surface for effects of quantity or

tension ... .. ... ... ... ... ... 83

Secondary battery of lead plates for prolonged effects of tension ... 84

Instructions respecting the use of secondary batteries ... ... 85

Returns. —Comparisons ... ... ... ... ... ... 88

SECOND PART.

APPLICATIONS.

Galvanocaustic applications ... ... ... ... ... 90

Use in lighting dark cavities in the human body, and obscure hollows in

general ... ... ... ... ... ... ... 94

Application to firing mines, etc. ... ... ... ... ... 95

Application to domestic uses.— Saturn's "tinder box " ... ... 97

Application to electric breaks for use on railways ... ... ... 102

Application to the eudiometric analysis of the atmosphere of mines ... 103

Application to the production of luminous signals .. ... .. 103

Application to the production of the electric light in special cases ... 104

Application to the sub-division of the electric light ... ... ... 105

Physiological effects produced by secondary batteries ... ... 106

Various applications ... ... ... ... ... ... 106

264

THIRD PART,

EFFECTS PRODUCED BY ELECTRIC CURRENTS OF HIGH TENSION.

CHAPTER I.

Page.

Use of secondary batteries for the study of these effects ... ... 108

Experiment upon the "luminous sheath" with the current decreasing in

intensity ... ... ... ... ... ... ... 110

Change of colour in the "luminous sheath" according to the intensity of

the current ... ... ... ... ... ... Ill

Batteries of from two hundred to eight hundred secondary cells, used in

studying electrical effects of high tension ... ... ... 113

Luminous liquid globules ... ... ... ... ... 117

Globular discharge. — Brush discharge and luminous figures produced by

the discharge of a battery of 800 secondary couples ... ... 120

Wandering electric sparks ... ... ... ... ... 123

Sheaf of aqueous globules ... ... ... ... ... 126

Jets of vapour ... ... ... ... ... ... ... 127

Electrified liquid vein — gyratory motion ... ... ... ... 128

Electric bar ... ... ... ... ... ... ... 130

Voltaic pump ... ... ... ... ... ... ... 131

Detonations produced at the extremity of the positive electrode ... 133

Electro-silicious light ... ... ... ... ... ... 135

Crowns, arcs, rays and undulating motions... ... ... ... 139

Electro dynamic whirls ... ... ... ... ... ... 140

Crater-like perforations ... ... ... .... ... ... 143

CHAPTER II.

Engraving on glass by elctricity ... ... ... ... ... 146

Electric boring or drilling ... ... ... ... ... 149

Different applications ... ... ... ... ... ... 150

265

FOURTH PART.

CHAPTER I.

Page.

Analogies with globular lightning... ... ... ... ... 151

On the formation of globular lightning .. ... ... ... 152

On the light emitted by globular lightning ... ... ... ... 154

On the noise accompanying globular lightning ... ... ... 156

On its gyratory motion ... ... ... ... ... ... 156

On its immediate disappearance ... ... ... ... ... 157

On the slow passage of globular lightning ... ... ... ... 157

On the noise accompanying the appearance and disappearance of globular

lightning ... ... ... ... ... ... ... 160

On the action of many pointed lightning conductors in cases of globular

lightning ... ... ... ... ... ... ' ... 162

Instance of globular lightning at Paris in 1876 ... ... ... 163

Lightning "en chapelet" ... ... ... ... ... 165

On the formation of lightning and its relations with globular lightning... 169

CHAPTER II.

Comparisons with the phenomenon of hail.

On the formation of hail ... ' ... ... ... ... 173

On the mechanical and calorific effects of electricity in the formation of

hail ... ... 174

On the wind, noise, lightning — with or without thunder — accompanying

hail ... ... ... ... ... ... 176

On hail produced without any apparent electrical manifestation ... 177

On the short duration of falls of hail ... ... ... ... 177

On the tracks of hail and rain .. ... .., ... ... 177

On the intermittences and returning forces of hail storms ... ... 178

On the form and glare of hailstones ... ... ... ... 179

On their internal structure and their size ... ... ... ... 179

On the whirlwinds accompanying hailstorms and the cause of their spiral

motion ... ... ... ... ... ... ... 181

Conclusions ... ... ... ... ... 182

266
CHAPTER III.

Comparisons with waterspouts,

Page.

On the luminous effects, noise, jets of vapour, &c., accompanying water-
spouts ... ... ... ... ... ... ... 183

On the spiral motion of whirlwinds and cyclones .,, ... ... 184

On the spiral motion of simoons ... ... ... ... ... 184

On the spiral motion of lightning... ... ... ... ... 186

On the waterspouts produced without apparent electrical phenomena ... 186

On the electrical signs of waterspouts .. ... ,., ... 186

On the cause of the descent of waterspouts ... ... ... 186

On the comparison of tidal waves and "bores" (seiches) with the

"electric bore" experiment ... ... ... ... ... 187

On the phenomena of the suction of waterspouts and their analogy with

the voltaic pump experiment ... ... ... ... ... 187

Conclusions ... ... ... ... ... 188

CHAPTER IV.

Comparisons with the polar aurora.

On the luminous effects, crowns and arcs and their undulating motion ... 189

On the dark segment or circle of the polar aurora ... ... ... 190

On the fluctuation of the light ... ... ... ... ... 191

On the falls of rain and snow, and the wind accompanying the polar

aurora ... ... ... ... ... .. ... 191

On the noise accompanying the polar aurora ... ... ... 192

On the magnetic disturbances ... ... ... ... ... 192

On the electrical signs of the polar aurora ... ... ... ... 192

On the cause of the discharge of electricity in the polar aurora ... 193

On the origin of atmospheric electricity ... ... ... ... 193

Conclusion ... ... ... ... ... ... ... 197

CHAPTER V.

Comparisons with spiral nebulse ... ... ... ... ... 198

267
CHAPTER VI.

Page.
Analogy with the solar spots ... ... ... ... 201

Upon the physical formation of the sun ... ... ... ... 203

FIFTH PART.

RHEOSTATIC MACHINE.

Description of the rheostatic machine ... ... ... ... 206

Effects produced by the rheostatic machine... ... ... ... 209

Crooked and "brush" form of the discharge ... ... ... 210

Light in a vacuum ... ... ... ... ... ... 212

On the possibility of obtaining some effects with a current of less tension 214

Complete transformation by means of the rheostatic machine of a certain

quantity of dynamic electricity stored by a secondary battery ... 214

Variation in the length of the spark in proportion to the number of con-
densers and to the tension of the current .. ... ... 215

Construction of a large rheostatic machine ... ... ... ... 216

Sparks produced under different conditions... ... ... ... 218

Form of the sparks produced by the large rheostatic machine... ... 220

Aborisations produced by the passage of the sparks ... ... ... 268

Lichtenberg figures produced by sparks from the rheostatic machine ... 224

Condensation of the sparks ... ... .. ... ... 229

Coloured sparks ... ... ... ... ... ... 230

Halo or red light in a vacuum ... ... ... ... ... 232

Means of valuing very small quantities of matter or very short intervals

of time by means of the rheostatic machine ... ... ... 233

Conclusion of the preceding observations ... ... ... ... 234

Effects of quantity from the rheostatic machine ... ... .. 235

Static sparks of quantity :.. ... ... ... ... 238

Observation on the noise produced by these sparks ... ... ... 240

Calorific effects... ... ... ... ... ... ... 240

Mechanical effects ... ... ... .. ... ... 241

Rheostatic hydraulic ram ... ... ... ... ... 241

268

Vibration knots formed in a metalllic wire by the current of quantity

from the rheostatic machine ... ... ... ... ... 242

Variation of the distance of the knots according to tension of the current 244

Noise in the wire ... ... ... 244

Fragility of the wire ... ... ... ... ... ... 246

Results relative to lightning conductors ... ... ... 246

Conclusions drawn from the preceding phenomena relative to the mode

of propagation of electricity ... ... ... ... ... 247

On the quality of reversal possessed by the rheostatic machine ... 248

SIXTH PART.

Analogies between electrical phenomena and effects produced
by mechanical actions.

Particular analogies between phenomena produced by currents of high

tension and the effects produced by mechanical actions ... ... 250

Analogies between different phenomena of static or dynamic electricity

and the effects produced by mechanical action ... ... ... 254

Conclusions relating to the nature of electricity ... ... ... 256

&R-A3&

~^se IHB

UHI7BRSIT7]

PARKER & HILL, PRINTERS, ALBERT STREET, BIRMINGHAM.

THIS BOOK IS DUE ON THE LAST DATE
STAMPED BELOW

AN INITIAL FINE OF 25 CENTS

WILL BE ASSESSED FOR FAILURE TO RETURN
THIS BOOK ON THE DATE DUE. THE PENALTY
WILL INCREASE TO 5O CENTS ON THE FOURTH
DAY AND TO $1.OO ON THE SEVENTH DAY
OVERDUE.

JAN 27 1936

947

Engineering
Library

••••••••••••I

mjjjjjjjgfjmjjj^jgmjjj^fjjjjji^jat

m

m

a.